Vision-Language Models — 2026 · Paper list (718)

July 2026 (308)

Date	Paper	Venue	Why selected
2026-07-23	K12-KGraph: A Curriculum-Aligned Knowledge Graph for Benchmarking and Training Educational LLMs	—	—
2026-07-23	Show, Don't Tell: Evaluating Spatial Cognition in Generative Pixels Rather Than LLM Text	—	—
2026-07-22	Test-Time Training for Modality Order Consistency in Vision-Language Models	—	Gandelsman (Berkeley); TTT fixes systematic VLM modality-order bias; reproducible across 3 models
2026-07-22	Look Less, Think Faster: Joint Token-Compute Adaptation for Multimodal LLMs	ECCV 2026	ECCV 2026; joint token-compute pruning cuts MLLM inference cost; directly deployable
2026-07-22	Trace: A Taxonomy-Guided Environment for Multidomain Visual Reasoning	—	RLVR training environment for VLMs with taxonomy-guided verifiable visual rewards
2026-07-22	LAVIFT: Latent-Action-Guided Vision Fine-Tuning for Surgical Interaction Recognition	—	Medical VLM fine-tuning for surgical instrument-tissue interaction; Intel Labs; reproducible
2026-07-22	ENTRAP-VL: A Taxonomic Probe for Dual Contextual Entrainment in Vision-Language Models	—	Taxonomic probe showing VLMs reliably succumb to irrelevant contextual entrainment
2026-07-22	SeededGrasp: Language-Guided Grasping in Complex Scenes with Multiple Embodiments	—	—
2026-07-22	ReferTrack: Referring Then Tracking for Embodied Visual Tracking	—	—
2026-07-22	Robostral Navigate	—	—
2026-07-21	Attributes Should Come from Images, Not Class Names: Distribution-Conditioned Attribute Selection for Vision-Language Models	—	Deepak Pathak + Zhiqiu Lin; LLM-generated CLIP descriptors carry no visual evidence — fundamental rethink
2026-07-21	Dual Adversarial Fine-tuning for Enhancing Robustness of Large Vision Language Model	—	ICT CAS (Shan/Chen/Gao); dual adversarial fine-tuning closes LVLM visual robustness gap
2026-07-21	Stochastic Meta-Unlearning: Bridging Language Backbone and Multimodal Unlearning	—	Sijia Liu + Tianlong Chen; first principled multimodal unlearning bridging language and vision components
2026-07-21	Mitigating Modality and Language-Style Gaps for Zero-Shot Video Moment Retrieval	ECCV 2026	ECCV 2026; zero-shot VMR resolving modality and language-style alignment gaps
2026-07-21	PathAgentBench: Benchmarking Evidence-Seeking Vision-Language Models on Whole-Slide Pathology Image	—	PathAgentBench; evidence-seeking VLMs evaluated on multi-scale whole-slide pathology
2026-07-21	ZeroSplat: Generalized Referring Segmentation in 3D Gaussian Splatting	ECCV 2026	ECCV 2026; multi-target referring segmentation extending 3DGS to language-guided scene understanding
2026-07-21	No Training, Better Flights: Test-Time Scaled VLMs for UAV Navigation	—	Test-time scaling lifts VLM navigation performance without any additional training
2026-07-21	Now You See the Hate: Adaptive View Retrieval for Hidden Hateful Illusions	—	VLM safety gap: state-of-the-art models under 25% accuracy on hateful optical illusions
2026-07-21	Fusion Embedding: A Unified Embedding Space for Text, Image, Video, and Audio	—	Unified text+image+video+audio embedding; fills critical audio gap in VLM retrieval stacks
2026-07-21	DobicVLM: Aligning Chest X-Ray Report Generation with Clinically-Grounded Programmatic Rewards via Group Relative Policy Optimization	—	GRPO-trained medical VLM for CXR reports; directly replicable RL recipe for domain-specific VLMs
2026-07-21	MeetingToM: Evaluating Multimodal LLMs on Theory-of-Mind Reasoning in Multi-Party Meetings	—	Multi-party ToM benchmark; measures social/epistemic reasoning gap in MLLMs for agentic settings
2026-07-21	ExpertVerse: A General-Purpose Benchmark for Expert-Level Reasoning in Knowledge-Intensive Visual Synthesis	—	Expert-level visual reasoning benchmark; stress-tests knowledge-intensive VLM synthesis beyond commonsense
2026-07-21	IGGT4D: Streaming 4D Instance-Grounded Geometry Transformer	—	Streaming 4D instance-grounded geometry; spatial intelligence backbone for embodied agents
2026-07-21	OPD-IAD: From Language Judgment to Industrial Anomaly Detection via On-Policy Self-Distillation	—	On-policy self-distillation adapts LVLM judgments to pixel-level IAD; novel spec-to-detector recipe
2026-07-21	HPD-Parsing: Hierarchical Parallel Document Parsing	—	Hierarchical parallel document parsing with VLMs; speed+accuracy gain for doc-understanding pipelines
2026-07-21	ABot-World-0: Infinite Interactive World Rollout on a Single Desktop GPU	—	—
2026-07-21	Computational Humor with Multimodal LLMs: Methods, Datasets, Evaluation, and Challenges	—	—
2026-07-21	DeforM: Reasoning-Guided Physics-Aware Video Generation via Spatial-Temporal Masking	—	Yu Qiao (Shanghai AI Lab); reasoning-guided physics-aware video generation via spatial-temporal masking
2026-07-21	ChronoStitch: Training-Free Composition of Visual KV Memories for Long-Horizon Temporal Reasoning	—	Training-free KV cache composition for long-video QA; avoids video reprocessing
2026-07-21	MissingBench-Verified: Probing Vision-Language Models' Inability to Detect Missing Object Parts	ECCV 2026	ECCV 2026; benchmark exposing VLM failure on missing-part detection; honest hallucination audit
2026-07-20	Patch Policy: Efficient Embodied Control via Dense Visual Representations	—	Yann LeCun co-author; dense ViT features for embodied control, underexplored direction
2026-07-20	The Many Senses of Visual Similarity: A Text-Prompted Image Perceptual Metric	—	Jun-Yan Zhu + Shechtman + Hertzmann; text-prompted multi-sense perceptual metric, novel
2026-07-20	Memory-Supported Synergistic Adaptation for Training-Free Test-Time Medical Image Segmentation	ECCV 2026	ECCV 2026; memory-synergistic TTA for medical VLM segmentation, classification→seg bridge
2026-07-20	O-VAD: Industrial Video Anomaly Detection through Object-Centric Tracking and Reasoning	ECCV 2026	ECCV 2026; object-centric VLM reasoning for industrial anomaly detection
2026-07-20	FlashRT: Agent Harness for Guiding Agents to Deploy Real-Time Multimodal Applications	—	Agent harness for real-time multimodal deployment; directly actionable for harness builders
2026-07-20	Anticipate Before Acting: Future-State-Conditioned Vision-Language Navigation	—	Future-state-conditioned VLN; novel supervision signal beyond next-action cloning
2026-07-20	Reasoning as a Double-Edged Sword: Architecture and Cross-Stage Robustness in Vision-Language-Action Models	—	Empirical robustness audit of reasoning in VLAs; honest baselines; safety-critical insight
2026-07-20	SEE: Structure-aware Exploring \& Exploiting for Long-horizon GUI Agent Trajectory Synthesis	—	GUI agent trajectory synthesis with VLMs; long-horizon coverage; agentic scaffolding
2026-07-20	Attention Without Grounding: Causal Evaluation of Visual Explanations in Medical VLMs	—	Causal audit proves attention heatmaps don't reflect actual medical VLM evidence
2026-07-20	Simple Domain Generalization for Strong Pixel-Level Image Tampering Detection in Modern VLMs	—	Domain-generalized pixel-level tampering detection in VLMs; practical safety/integrity tooling
2026-07-20	FlowMimic: Mask-free Visual Editing and Generation with Pixel-pair Warped Flow Field for Online Video Editing Data Generation and Modality Mimicry	—	—
2026-07-20	PRiSM: Prototype Regularization for Few-Shot VLMs	—	Dolz + Ben Ayed; few-shot VLM adaptation under realistic class-imbalance; buildable recipe
2026-07-20	Measuring and Improving Complex-Atomic Answer Consistency in Endoscopic VQA	—	Complex-atomic answer consistency metric for endoscopic VQA; fills medical eval gap
2026-07-19	TimeLens2: Generalist Video Temporal Grounding with Multimodal LLMs	—	Generalist video temporal grounding with MLLMs; fills key capability gap in video VLMs
2026-07-19	EvoGUI: An Evolution-Aware Benchmark for GUI State-Transition Understanding	—	Jiwen Lu (Tsinghua); EvoGUI isolates state-transition reasoning from end-to-end GUI success
2026-07-19	Searching for Task-Specific Vision Paths: Evolutionary Block Pruning Across Vision-Language Models	—	Evolutionary block pruning finds task-specific vision paths; cuts VLM compute without retraining
2026-07-18	Look Clearly Before Answering: Mitigating Hallucinations in LVLMs via Saliency-Driven Perceptual Realignment	—	Saliency-driven hallucination mitigation in LVLMs; core reliability problem, buildable method
2026-07-18	Dataset Distillation by Influence Matching	CVPR 2026	CVPR 2026; outcome-centric dataset distillation; broadly applicable to VLM training data
2026-07-18	Can Experts Adapt Without Training? On Test-Time Modality Generalization in MVLMs	MICCAI 2026	MICCAI 2026; test-time modality generalization for medical VLMs; directly buildable
2026-07-18	HTT-Net: Hierarchical Text-guided Transition Modeling for Surgical Video Phase Recognition	—	Text-guided surgical video phase recognition; strong medical VLM baseline with temporal structure
2026-07-17	Ask Twice, Look Twice: Prompt Echoing Resolves the Question-First Paradox in Vision-Language Models	—	Deva Ramanan (CMU); prompt-position paradox fix; universal VLM prompting impact
2026-07-17	Von Mises-Fisher Mixture Model with Dynamic Shrinkage for Realistic Test-Time Transduction	ICML 2026	ICML 2026; vMF mixture TTA-transduction; rigorous test-time VLM adaptation theory
2026-07-17	More with Less: a Large Scale Remote Sensing VLM with a Simple Recipe	—	Luc Van Gool (ETH); simple recipe beats RS-specific archs; transferable scaling insight
2026-07-17	Model Merging for Medical LVLMs: A Benchmark and a Winner-Take-All Approach	—	Medical LVLM model-merging benchmark + WTA method; practical LoRA-expert fusion
2026-07-17	How Do VLMs Fail? Vision-Operation Misalignment in Compositional VQA	—	Vision-operation misalignment taxonomy; actionable failure map for compositional VQA
2026-07-17	ToolSciVer: Multimodal Scientific Claim Verification with Visual Tool Augmented Reinforcement Learning	—	VL tool-use + RL for multimodal scientific verification; novel reward signal design
2026-07-17	When Can Test-Time Adaptation Help Zero-Shot CT Vision-Language Models?	—	Leonid Sigal (UBC); systematic TTA scope study on zero-shot 3D CT VLMs
2026-07-17	Region-Grounded Vision-Language Learning for Detection-Guided Mammographic Lesion Classification	—	Region-grounded contrastive VL for mammography; spatial alignment beats global CLIP
2026-07-17	Knowing the Self, Understanding the World: A Dual-Cognition Benchmark for UAV Spatio-temporal Reasoning with MLLMs	—	—
2026-07-17	Vision-Language Assistant for Emotional Reactions to Risky Driving	—	—
2026-07-17	An Exam for Active Observers	—	—
2026-07-17	Attention-Guided Saliency Maps for Interpreting Visualization Literacy in VLMs	—	—
2026-07-17	Exo2EgoPose: Leveraging Exocentric Demonstrations for Vision-Language guided Egocentric 3D Hand Pose Forecasting	—	—
2026-07-17	Debiasing Text-to-Image Evaluation via Implicit Cultural Alignment Reward Modeling	—	—
2026-07-17	One Modality to Forget Them All: Enhancing Cross-Modal Unlearning in Vision-Language Models	—	Cross-modal unlearning vulnerability in VLMs; important safety gap with concrete attack surface
2026-07-17	Foresight Residual RL for Long-Horizon Robot Manipulation with Vision-Language-Action Models	—	VLA + residual RL for long-horizon manipulation; Boularias lab; addresses credit assignment gap
2026-07-17	JoyNexus: Service-Oriented Multi-Tenant Post-Training for VLA Models	—	Multi-tenant post-training service for VLA; novel deployment angle; Wentao Zhang group
2026-07-17	PhysAgent: Reflective Agentic Physics Control for Physically Plausible Video Generation	—	Reflective agentic physics control for video gen; VLM+agent+simulation pipeline
2026-07-16	Symbal: Detecting Systematic Misalignments in Model-Generated Captions	ICML 2026	Stanford/Langlotz group; ICML 2026; systematic caption error detection for medical MLLMs
2026-07-16	Multi-LLM Collaborative MRI Report Generation for Visual Instruction Tuning in Brain Oncology	—	Multi-LLM VLM pipeline for 3D brain MRI reports; novel visual instruction tuning for oncology
2026-07-16	Action QFormer: Structured Representation Shaping under Action Supervision in Vision-Language-Action Models	—	Action supervision reshapes multimodal representations in VLA; structural insight for builders
2026-07-16	FoMoVLA: Bridging Visual Foresight and Motion Guidance for Vision-Language-Action Models	—	Visual foresight + motion guidance in VLA models; explicit forward prediction addresses reactivity
2026-07-16	WorkDrive: Roadwork Chain of Causation for Autonomous Driving	—	Chain-of-causation reasoning for roadwork zones; tackles hard OOD failure mode in driving VLMs
2026-07-16	HoloGeo: Mitigating Landmark Bias in Geo-localization via Evidence-Driven Reasoning	—	Evidence-driven reasoning to mitigate landmark bias in VLM geo-localization
2026-07-16	RoboTTT: Context Scaling for Robot Policies	—	—
2026-07-16	SportD: Can VLMs Physically Strategize?	—	Probes VLM physical strategy/reasoning in soccer; novel eval axis for agent decision-making
2026-07-16	U-shaped Multi-granularity Learning for Vision-Language Models	—	Prompt learning granularity dilemma for VLMs; practical recipe for cross-task generalization
2026-07-16	GeoDetect: Geometric Adversarial Detection for VLPs	ECCV 2026	Adversarial detection for multimodal VLPs; ECCV 2026; geometric approach is novel angle
2026-07-16	VTM-Nav: Hierarchical Visual-Topological Memory for Cross-Episode Object-Goal Navigation	—	VLM-driven object-goal navigation with topological memory; training-free embodied agent system
2026-07-16	On Success and Simplicity: A Second Look at Transferable Vision-Language Attack Pipeline	—	Simplicity-first transferable attack on VLPMs; questions complexity of existing pipelines
2026-07-16	ARMOR++: Agentic Orchestration of a Multi-Domain Primitive Set for Transferable Attacks on Deepfake Detectors	—	—
2026-07-16	Knowing You at First Glance: Inferring Apparent Personality from Faces	—	—
2026-07-16	Xiaomi-Robotics-1: Scaling Vision-Language-Action Models with over 100K Hours of Real-World Trajectories	—	Xiaomi robotics; 100K hrs real VLA trajectories; massive-scale grounded VLM-action
2026-07-16	Video = World + Event Stream	—	Wan-Streamer (Alibaba); world+event decomposition reframes video VLM architecture
2026-07-16	Training-Free Open-Vocabulary 3D Point-Cloud Segmentation on the Generalized Few-Shot Benchmark	—	—
2026-07-15	Exploratory, Communicative, and Deployable: Vision-Driven Embodied Agents for Open-World Mobile Manipulation	ECCV 2026	—
2026-07-15	ScanFocus: A Coarse-to-Fine Framework for Spatio-Temporal Video Grounding	ECCV 2026	—
2026-07-15	M\(^\text{4}\)World: A Multi-view Multimodal Driving World Model for Interactive Object Manipulation and Minute-long Streaming	—	—
2026-07-15	SIVA-RL: Sensitivity-Invariance Visual Alignment for Multimodal Reinforcement Learning	—	—
2026-07-15	Fine-Grained Vision-Language Pretraining with Organ-Conditioned Pattern Tokens for CT Understanding	—	—
2026-07-15	Groc-PO: Grounded Context Preference Optimization for Truthful Multimodal LLMs	—	—
2026-07-15	Fine-grained CLIP fine-tuning with self-annotated region alignment	—	—
2026-07-15	Attention-Free and Lightweight Token Reduction for Efficient Vision-Language Models	—	Attention-free token reduction; directly enables VLM deployment on edge/resource-constrained devices
2026-07-15	Generalizable VLA Finetuning via Representation Anchoring and Language-Action Alignment	—	Representation anchoring + language-action alignment; fixes BC fine-tuning catastrophic forgetting in VLAs
2026-07-15	Semantic Anchoring for Robotic Action Representations	—	Semantic anchoring preserves VLM priors in VLA fine-tuning; Yizhou Wang (PKU) group
2026-07-15	Self-Improving is Often Sudden: Enlightenment-style Finetuning for Large-Scale Models	—	Phase-transition self-improvement phenomenon; Tianwei Zhang (NTU); actionable finetuning recipe
2026-07-15	FM\(^2\): Unified Federated Foundation Models for Heterogeneous Multimodal Medical Imaging	—	Federated multimodal medical VLM; addresses privacy-constrained cross-institution medical AI deployment
2026-07-15	SD-MAR: Multi-image Analytical Reasoning via Synthetic Data and Reinforcement Learning	—	RL + synthetic data for multi-image analytical reasoning; addresses core VLM reasoning gap
2026-07-15	MonteRET: AI Agent Enhancing Multimodal LLMs with Multi-granularity Knowledge Retrieval for Chest CT Report Generation	—	Multi-granularity RAG agent for chest CT reports; retrospective clinical evaluation included
2026-07-15	DiMaS: Distribution Matching for Steering Vision-Language-Action Models	—	Fine-grained behavioral steering for flow-matching VLA; enables controllable robot policy
2026-07-15	Never Too Late for Force: Accelerating VLA Post-Training with Reactive Force Injection	—	Force injection during VLA post-training; addresses contact-state blindspot in vision-driven policies
2026-07-15	RxBrain: Embodied Cognition Foundation Model with Joint Language-Visual Reasoning and Imagination	—	—
2026-07-15	GHR-VLM: Making Zero-Shot Transit Video Analytics Realizable with Grounded Hybrid Reasoning	—	Zero-shot transit video analytics with grounded hybrid VLM reasoning; practical deployment recipe
2026-07-15	Multi-Head Latent Control: A Unified Interface for LLM Agent Decision Making	—	—
2026-07-15	ThinkBLOX: 3D Indoor Scene Generation with Progressive Reasoning	—	—
2026-07-14	VisCo: Leveraging Large Language Models as Intrinsic Encoders for Visual Token Compression	—	Visual token compression for VLMs; training-free LLM-as-encoder approach; addresses core inference efficiency bottleneck
2026-07-14	CoRe: A Comprehensive Framework for Cross-Image Comparative Reasoning in Vision-Language Models	—	CoRe framework for cross-image comparative reasoning; fine-grained attribute grounding; fills known VLM compositional gap
2026-07-14	Gaussian Mixture Modeling for Event-Aware Visual Allocation in Long Video Understanding	—	Gaussian mixture keyframe selection for long video VLMs; event-aware visual allocation; addresses uniform-sampling information loss
2026-07-14	Hy-Embodied-VLM-1.0: Efficient Physical-World Agents	—	Hy-Embodied-VLM-1.0; end-to-end embodied agent with multimodal perception and agentic reasoning; buildable system report
2026-07-14	Auditing Data Leakage in Whole-Slide Image Multimodal Benchmarks	—	Data leakage audit in WSI pathology VLM benchmarks; questions benchmark validity; critical for medical VLM evaluation
2026-07-14	A Learning-Rate-Gated Failure of GRPO in a Small Language and Vision-Language Model Web Agent: A Controlled Null and Its Mechanism	—	Controlled null result on GRPO failing for small VLMs; mechanistic analysis; essential calibration for agent RL fine-tuning
2026-07-14	What Does a Temporal Benchmark Score Measure? Decomposing Channel Use in Video VLM Evaluation	—	Decomposes what temporal benchmark scores actually measure in video VLMs; channel-use analysis; methodological contribution
2026-07-14	Let RGB Be the Language of Vision	—	Unified RGB formulation for all visual modalities (masks, depth, etc.); architectural simplification with broad VLM implications
2026-07-14	ReflectVLN: Training Vision-Language Navigation Agents with Reflective Reasoning	—	ReflectVLN adds closed-loop reflective reasoning to VLN agents; explicit failure diagnosis mechanism; directly applicable to agent harnesses
2026-07-14	The Sound of Absence: Audio-Language Embedding Models Struggle with Negation	—	Hung-yi Lee (NTU); exposes fundamental negation blindspot in CLAP-style audio-language models
2026-07-14	Navigating the Mirage: A Dual-Path Agentic Framework for Robust Misleading Chart Question Answering	—	—
2026-07-14	MQAdapter: Multi-Modal Quantum Adapter for Coarse-to-Fine VLM Fine-tuning	—	Quantum-inspired adapter for coarse-to-fine VLM fine-tuning; novel PEFT direction
2026-07-14	DM-KG: A Novel Method for Boosting Spatial Cognition of Vision-Language Models in Street View Imagery	—	Spatial cognition augmentation for VLMs on street-view; directly relevant to geo-VLM eval
2026-07-14	Lost in Visual Translation: A VLM-Assisted Perceptual-Semantic Coherence Framework for EEG-to-Image Reconstruction	—	VLM-assisted perceptual-semantic coherence for EEG-to-image; novel eval framework
2026-07-14	Breaking Déjà Vu: Independent Auditing of Visual Place Recognition through Vision-Language Reasoning	—	VLM reasoning for visual place recognition auditing; novel robustness eval for embodied nav
2026-07-14	Text2Sign: A Single-GPU Diffusion Baseline for Text-to-Sign Language Video Generation	—	—
2026-07-13	MED-DSLC: Multi-Expert-Domain Classification via Domain Supervision and Logit Calibration	ECCV 2026	ECCV 2026; Nuno Vasconcelos (UCSD); fixes CLIP logit non-comparability across domains—principled zero-shot fix
2026-07-13	StructAgent: Harness Long-horizon Digital Agents with Unified Causal Structure	—	Causal structure for long-horizon VLM digital agents; directly relevant to agent harness design
2026-07-13	Confidence Scores in Open-Vocabulary Detection Are a Biased Mixture of Scale and Semantics	—	Open-vocab detector confidence = scale+semantics mixture; exposes fundamental reliability flaw in CLIP-based OVD
2026-07-13	TIGER: Text-Conditioned Visual Gated Routing with Acceptance Alignment for Multimodal Speculative Decoding	—	Multimodal speculative decoding with gated routing; practical inference speedup for VLMs
2026-07-13	When Depth Is Better Told Than Shown: Depth-Ordinal Prompting for Vision-Language Spatial Reasoning	—	Depth-ordinal prompting for VLM spatial reasoning; addresses a core 3D perception weakness in VLMs
2026-07-13	Think When It Matters: Conditional VLM Reasoning for Social Navigation with RL Policies	—	Conditional VLM reasoning with RL for social navigation; conditional compute in embodied agents
2026-07-13	StoryTeller: Training-Free Narrative Grounding for Long-Form Audio Description	ECCV	—
2026-07-13	DynEval: Holistic Evaluations of T2I Generative Models in the Wild	ECCV 2026	—
2026-07-13	An Empirical Analysis of Continual Learning for Heterogeneous Medical Visual Question Answering	—	Empirical continual-learning study for medical VQA; practical for multi-task clinical VLM deployment
2026-07-13	LaGuadia: Language-Guided Adaptive Distillation from Pathology Foundation Models	—	Language-guided distillation from pathology FMs; Won-Ki Jeong; reduces WSI inference cost
2026-07-13	MAGIC: Transition-Aware Generation of Navigable Multi-Scene Game Worlds with Large Language Models	—	—
2026-07-13	The Emerging Paradigm of Geospatial Foundation Models: From Pre-Training to Agentic Reasoning	—	—
2026-07-13	SVR-R1: Bootstrapping Multi-modal Reasoning with Self-verification in Reinforcement Learning	—	Self-verification RL for multimodal reasoning; extends R1-style training to VLMs
2026-07-13	See like a Robot: Robot-Centric Pointmaps for Vision-Language-Action Models	—	Robot-frame 3D pointmaps for VLA; fixes camera-frame mismatch, strong geometric inductive bias
2026-07-12	Spectral Heat Flow for Conservative Token Condensation in Vision-Language Models	ICML 2026	ICML 2026; spectral-heat-flow token condensation; directly cuts VLM inference cost without info loss
2026-07-12	Learning Anatomy-Grounded CT Vision-Language Representations with Organ-Hierarchical Report Knowledge	—	Anatomy-hierarchical VLP from CT+reports; strong medical VLM pretraining recipe for organ-level grounding
2026-07-12	3D-DefectBench: A Controlled Factorial Study of Vision-Language Model Evaluation Pipelines for Fine-Grained 3D Generation Defects	—	Factorial study of VLM eval pipelines for 3D generation; rigorous methodology for automated judge reliability
2026-07-12	Mixture of Cognitive Experts in Large Vision-Language Models	—	Mixture-of-cognitive-experts for LVLMs; novel framing of metacognition and diverse representations
2026-07-12	Detecting AI-Generated Video: A Vision-Language Dual-View Survey	ACL 2026	—
2026-07-12	Traj-VLN: Learning Pixel-Space Interaction via Autoregressive Trajectory Generation	—	Autoregressive pixel-space trajectory generation for VLN; novel grounding approach for navigation agents
2026-07-11	What Does Your Short-Answer VQA Score Actually Measure? Evaluator-Dependent Instability in Multimodal Short-Answer Benchmarks	—	Demonstrates VQA benchmark scores are evaluator-dependent; critical for anyone running VLM evals
2026-07-11	Empowering Long-form Omni-modal Understanding with Robust Audio Perception	—	Weidi Xie (SJTU/Oxford); long-form omni-modal understanding with explicit audio alignment; scarce labeled data angle
2026-07-11	SynthDocBench: Controlled Benchmark for Long-Context Visual Document Understanding	COLM 2026	COLM 2026; controlled benchmark for long-context VLM document understanding, fills real eval gap
2026-07-11	PhysMRV: Physical Memory Retrieval and Verification for Physics Plausibility Reasoning	—	Physical plausibility reasoning in video-LMs; targets a known VLM failure mode with retrieval+verification
2026-07-11	Benchmarking the Robustness of Foundation Models for Mammography under Domain Shift	—	Benchmarks 15 foundation model backbones for mammography under domain shift; directly relevant to medical VLM eval
2026-07-11	WeaveEarth: Structured Evidence Construction and Reasoning for Training-Free UHR Remote Sensing Understanding	—	Training-free UHR remote sensing VLM reasoning with structured evidence; novel efficiency angle
2026-07-11	ABot-AgentOS: A General Robotic Agent OS with Lifelong Multi-modal Memory	—	Robotic agent OS with lifelong multimodal memory; directly relevant to agent harness architecture
2026-07-11	Imperceptible and Reversible Adversarial Examples against Vision-Language Models for Privacy Protection	—	Reversible adversarial examples against VLMs for privacy; novel threat model with practical reversibility
2026-07-11	ABot-N1: Toward a General Visual Language Navigation Foundation Model	—	Foundation model for Visual Language Navigation; unifies spatial reasoning and embodied versatility
2026-07-10	ALICE: Learning a General-Purpose Pathology Foundation Model from Vision, Vision-Language, and Slide-Level Experts	—	Pathology foundation model fusing vision, VL, and slide-level experts; builds on strong computational pathology trend
2026-07-10	Video Generation Models are General-Purpose Vision Learners	ECCV 2026	Video generation as general vision learner; ECCV 2026; Jasper Uijlings (Google); strong foundational claim
2026-07-10	SigLIP-HD by Fine-to-Coarse Supervision	ICLR 2026	SigLIP-HD fine-to-coarse supervision; ICLR 2026; Hengshuang Zhao group; directly improves MLLM visual encoding
2026-07-10	Super-Generalist: Towards Comprehensive and Accurate Medical Image Understanding via Generalist-Specialist Synergy	—	Generalist-specialist synergy for medical image understanding; addresses known VLM accuracy gap in clinical settings
2026-07-10	MOSAIC: Adaptive Inter-layer Composition for Efficient Heterogeneous Vision-Language Models	—	Heterogeneous VLM with linear attention interleaving; efficiency + performance; actionable architecture direction
2026-07-10	Robustifying Vision-Language Models via Test-Time Prompt Adaptation	ICML 2026	ICML 2026; test-time prompt adaptation for adversarial robustness in VLMs; strong venue, practical defense method
2026-07-10	The Count Is There, but Misaligned: Understanding and Correcting Counting Failures in VLMs	—	Diagnoses counting failures as representation-verbalization gap; mechanistic insight actionable for VLM evaluation
2026-07-10	Test-Time Scaling for Small VLMs on Multilingual Visual MCQ	—	Test-time scaling on small open VLMs—transfers TTS gains beyond frontier models; highly actionable
2026-07-10	Seeing is Free, Speaking is Not: Uncovering the True Energy Bottleneck in Edge VLM Inference	—	Reveals decode (LLM text gen) not vision tokens as edge energy bottleneck; inverts prevailing optimization target
2026-07-10	Evolution of Accuracy and Visual-Cognitive Errors in a Decade of Vision-Language AI Models	—	Decade-long longitudinal analysis of VLM visual-cognitive errors; rare rigorous error taxonomy work
2026-07-10	TCLA: Training-Free Class-wise Logit Adaptation for Medical Vision-Language Models	—	Training-free class-wise logit adaptation for OOD medical VLMs; lightweight, reproducible, practical
2026-07-09	Dive Into the Implicit Biases of Low-rank Vision-language Alignment	ECCV 2026	ECCV 2026; challenges full-param alignment dogma with low-rank analysis — rethinks VLM training recipe
2026-07-09	Cognitive-structured Multimodal Agent for Multimodal Understanding, Generation, and Editing	—	Cognitive memory architecture for unified understand/generate/edit — directly improves multimodal agent design
2026-07-09	When Thinking Hurts: Epistemic Signals in the Reasoning Chains of Visual Language Models	ICML 2026	ICML 2026; first empirical characterization of reasoning-chain uncertainty in thinking-mode VLMs — calibration insight
2026-07-09	APIVOT: Adaptive Planning with Interleaved Vision-Language Thoughts	—	Stanford (Haber lab); interleaved vision-language thoughts for long-horizon robot planning — actionable VLA scaffold
2026-07-09	Rethinking Small VLM Quantization: From Component-Wise Analysis to Hardware-Aware Edge Deployment	ICML 2026	ICML 2026; component-wise quantization analysis for sub-3B VLMs — direct edge deployment recipe
2026-07-09	Playing ZendoWorld: Challenging AI Agents on Active Visual Concept Induction	—	ZendoWorld benchmark for active visual concept induction — novel evaluation of hypothesis-driven VLM reasoning
2026-07-09	When Structured Sparse Autoencoders Learn Consistent Concepts Across Modalities	—	Sparse autoencoders for cross-modal mechanistic interpretability in VLMs — novel probe into internal representations
2026-07-09	VEGAS: Human-Aligned Video Caption Evaluation via Gaze	—	Gaze-grounded VLM caption eval; training-free metric; novel human-attention alignment signal
2026-07-09	LEEVLA: Seeing What Matters in Latent Environment Evolution for Vision-Language-Action	—	Selective visual token routing in VLA; addresses core dynamic-scene limitation in robot VLMs
2026-07-09	AUTOPILOT VQA: Benchmarking Vision-Language Models for Incident-Centric Dashcam Understanding	CVPR	CVPR dashcam VQA benchmark; incident-centric eval fills gap in autonomous driving VLM testing
2026-07-09	Post-Training in End-to-End Autonomous Driving	—	Comprehensive post-training survey for E2E autonomous driving VLAs; strong practical recipes
2026-07-09	WCog-VLA: A Dual-Level World-Cognitive Vision-Language-Action Model for End-to-End Autonomous Driving	—	Dual-level world-cognitive VLA for autonomous driving; addresses reactive limitation with world foresight
2026-07-09	VocaDet: Sample-Driven Open-Vocabulary Object Detection and Segmentation via Visual Tokenization and Vector Database Retrieval	—	Visual tokenization + vector DB for open-vocab detection/segmentation; novel retrieval-grounded approach
2026-07-09	Stop Guessing When to Stop Testing: Efficient Model Evaluation with Just Enough Data	—	Adaptive benchmark sizing for model evaluation; directly actionable for rigorous VLM evaluation practice
2026-07-09	Open-ended Multi-agent Autocurricula via Visual Inspection of Policies with Multi-modal LLMs	—	Multimodal LLMs as policy inspectors for RL curricula; novel VLM-as-evaluator application
2026-07-09	Attribute Retrieving for Open-Vocabulary Endoscopic Compositional Referring Segmentation	—	Attribute-driven compositional referring segmentation for endoscopy; medical VLM with fine-grained control
2026-07-09	Write-Protected Discrete Bottlenecks for Language-Grounded World Models: A Structural Limitation and Sufficient Fix	—	Structural analysis of language-grounded world models; critiques RT-2/PaLM-E paradigm with fix
2026-07-09	MultiView-Bench: A Diagnostic Benchmark for World-Centric Multi-View Integration in VLMs	—	Multi-view integration benchmark; targets untested allocentric 3D reasoning gap in VLMs; diagnostic benchmark value
2026-07-09	Mixture of Probes: Learning from Privileged Modalities in Multimodal LLMs Through Probing	—	Learning from privileged modalities via probing; addresses inference-time modality mismatch; practical for deployment
2026-07-09	Blind-Spots-Bench: Evaluating Blind Spots in Multimodal Models	—	Blind-Spots-Bench; systematic evaluation of multimodal model failure modes humans find trivial; benchmark for VLM robustness
2026-07-08	HIVE: Understanding Post-Hallucination Reasoning in Vision Language Models	ECCV 2026	ECCV 2026; dissects post-hallucination reasoning chains in VLMs, not just detection
2026-07-08	AnchorPrune: Relevance-Anchored Contextual Expansion for Visual Token Pruning	ECCV 2026	ECCV 2026; query-relevance-anchored visual token pruning cuts VLM inference cost
2026-07-08	Dual Latent Memory in Vision-Language-Action Models for Robotic Manipulation	—	—
2026-07-08	Comparative Study of Domain-adapted VLMs for General Document Visual Question Answering	—	Comparative domain-adapted VLMs for DocVQA — honest baselines across architectures for document understanding
2026-07-08	When Prompts Ignore Structure: Graph-Based Attribute Reasoning for Calibrated VLMs	—	Graph-based attribute reasoning for VLM calibration — addresses overconfidence in prompt-tuned models
2026-07-08	UP: Unbounded Positive Asymmetric Optimization for Breaking the Exploration-Stability Dilemma	—	—
2026-07-08	Multimodal Unlearning Across Vision, Language, Video, and Audio: Survey of Methods, Datasets, and Benchmarks	ACL	ACL survey; comprehensive multimodal unlearning across modalities; timely safety/privacy topic
2026-07-08	Evaluation of Multilingual Ability to Use Spatial Deictic Expressions in Vision-Language Models	ACL	ACL; multilingual spatial deictic eval; exposes systematic VLM spatial reasoning failure mode
2026-07-08	On Adversarial Vulnerability of Vision-Language Models through the Lens of Intermediate Spectral Subspaces	—	Spectral subspace lens on VLM adversarial vulnerability; novel mechanistic robustness analysis
2026-07-08	InfraQR: Edge-Placed QR-Inspired Structured Patch Attacks on Infrared Vision-Language Models	—	Structured adversarial patch attacks on infrared VLMs; underexplored robustness axis with security relevance
2026-07-08	Generalist Vision-Language Models for Fast Radio Burst detection: a zero-shot benchmark against a specialized detector	—	Zero-shot VLM benchmark vs. specialized FRB detector; honest generalist vs. specialist evaluation
2026-07-08	Heterogeneity-Adaptive Diffusion Schrodinger Bridge for PET-Guided Whole-Body MRI Translation	MICCAI 2026	—
2026-07-08	Prototype-Anchored Generalized Manifold Regression for Unknown-Domain Object Detection	—	—
2026-07-08	BUS: Brain-Inspired Unsupervised Self-Reflection via Backward Prediction for Multimodal Reasoning	—	Brain-inspired backward prediction for VLM self-reflection; novel training-free reasoning improvement mechanism
2026-07-07	Analysis-by-Proxy: Localization Signals in VLMs Operating as Condition Encoders	ICML 2026	ICML 2026; Daniel Cohen-Or/Or Patashnik; novel mechanistic insight into VLM localization signals
2026-07-07	Bridging Physical Reasoning and Task Generalization via Visual Action Outcome Reasoning Alignment	ICML 26	ICML 2026; VLM physical reasoning failure modes directly impact VLA agent generalization
2026-07-07	SAMPLe: SAM-based Optimizer for Prompt Learning in VLMs	ECCV	ECCV; SAM-based optimizer for VLM prompt tuning; directly addresses performance-generalization tradeoff
2026-07-07	What Images Cannot Say: Language-Guided Olfactory Representation Learning	ECCV 2026	—
2026-07-07	Lift3D-VLA: Lifting VLA Models to 3D Geometry and Dynamics-Aware Manipulation	—	—
2026-07-07	MonoIR-RS: Infrared Remote Sensing Vision-Language Learning with CLIP and VLM Adaptation	—	—
2026-07-07	AirflowAttack: Thermal-Airflow Adversarial Perturbations against Infrared Remote-Sensing Vision-Language Models	—	—
2026-07-07	A VLM-Enhanced Framework for Comprehensive Traffic Sign Condition Assessment Integrating Daytime Visual Performance and Nighttime Retroreflectivity Evaluation	—	—
2026-07-07	VaseMuseum: Digital Intelligent Museum for Ancient Greek Pottery	—	—
2026-07-07	Training-Free Acceleration for Vision-Language-Action Models with Action Caching and Refinement	—	—
2026-07-07	TMF-RSE: Tri-Modal Fusion with Regional Semantics and Evidential Uncertainty for Lung Severity Scoring	—	—
2026-07-07	PolicyShiftGuard: Benchmarking and Improving Policy-Adaptive Image Guardrails	—	Policy-adaptive image guardrails benchmark; critical for real deployment safety
2026-07-07	Pluralis v0.1: Towards a Multicultural, Multimodal, Multilingual Benchmark for AI Risk and Reliability	—	Multicultural/multilingual VLM safety benchmark exposing Western-centric evaluation gaps
2026-07-07	Nemotron-Labs-Diffusion: A Tri-Mode Language Model Unifying Autoregressive, Diffusion, and Self-Speculation Decoding	—	—
2026-07-07	SIEVE: Structure-Aware Data Selection for Imitation Learning with VLA Models	—	—
2026-07-07	RoboTALES: Learning Reasoning-Guided Robot Policies via Task-Aligned Simulated Futures	—	—
2026-07-07	Structured-Condensed Prompt Tuning in Vision-Language Models for Fine-grained Image Recognition	—	Structured prompt tuning for CLIP fine-grained recognition — practical, reproducible PEFT recipe for VLMs
2026-07-07	Reward Valuation in Vision Language Models: Causal Mechanisms Underlying Anhedonia	—	Schrimpf lab; mechanistic VLM–cognition alignment; novel anhedonia/reward valuation probe
2026-07-07	Pelican-VLA 0.5: Attending Before Acting Benefits Generalization	—	Unified VLA with attention-before-action; integrates future-frame gen + action in one model
2026-07-07	UI2App: Benchmarking Visual Interaction Inference in Executable Web Application Generation	—	UI-to-app generation benchmark; visual interaction inference; novel multimodal code-gen eval
2026-07-07	PVCap: Towards Accurate 3D Dense Captioning via PseudoCap and VoxelCapNet	—	3D dense captioning with pseudo-cap + voxel net; concrete architectural advance for 3D VLMs
2026-07-07	Vision Language Action (VLA) Models for Unmanned Aerial Robotics and Bimanual Manipulation: A Review	—	Review of VLA models across UAV and bimanual manipulation; useful scope survey for builders
2026-07-07	VendorBench-100: A Unified Cross-Paradigm Benchmark for Deepfake Image Detection	—	—
2026-07-06	TORINO: Token Reduction via Interpretable Concept Overlap in Vision-Language Models	—	Interpretable concept-overlap token reduction cuts VLM compute with explainable mechanism
2026-07-06	Does It Fail to See or Fail to Know? Attributing Errors in Vision-Language Models	—	Disentangles perception vs. knowledge failures in VLM errors — useful for evaluation design
2026-07-06	Solve the Missing First Step: Can VLMs Standardize Raw Heterogeneous Medical Data?	—	VLMs for medical data standardization — tackles missing preprocessing step in clinical AI pipelines
2026-07-06	QSVideo: Query-Conditioned Semantic Temporal Retrieval for Video Understanding	ECCV 2026	ECCV 2026; query-conditioned temporal retrieval tackles core long-video VLM degradation problem
2026-07-06	PixelPilot: Scalable Vision-Language-Action Models for End-to-End Autonomous Driving	ECCV 2026	ECCV 2026; end-to-end VLA for autonomous driving; scalable architecture with practical evaluation
2026-07-06	InternVLA-A1.5: Unifying Understanding, Latent Foresight, and Action for Compositional Generalization	—	InternVLA series (Shanghai AI Lab); unifies semantic priors and future prediction in one VLA
2026-07-06	Cortex: A Bidirectionally Aligned Embodied Agent Framework for Long-horizon Manipulation	—	Bidirectional hierarchical VLA agent addressing Markovian failure in long-horizon manipulation
2026-07-06	Do All Visual Tokens Matter Equally? Object-Evidence Preserving Token Merging for Vision-Language Retrieval	—	Object-evidence token merging for dense retrieval; practical efficiency for deploying VLMs
2026-07-06	Is the Geometry Doing the Work? An Operating-Point Audit of Hierarchy in Hyperbolic Vision-Language Models	TMLR	—
2026-07-06	StructuredEdit: Constraint-Aware Graphic Design Editing via Differentiable Parameter Propagation	SIGGRAPH	—
2026-07-06	DGSeg: Dynamic Gating of Semantic-Spatial Guided Predictions for Reasoning Segmentation	—	Dynamic gating for reasoning segmentation bridges language queries to pixel masks
2026-07-06	Repurposing CLIP to Localize at Pixel Level	—	Repurposes CLIP for pixel-level localization, addressing global-feature bias bottleneck
2026-07-06	Beyond Isolated Objects: Relationship-aware Open Vocabulary Scene Understanding via 3D Scene Graph Analysis	—	3D scene graph + VLM open-vocab scene understanding beyond isolated object lifting
2026-07-06	HunyuanOCR-1.5: Making Lightweight OCR VLMs Faster and Better	—	Tencent HunyuanOCR-1.5; lightweight OCR-VLM unifying doc parsing, extraction, translation
2026-07-05	DynaVieW: Schema-Guided World Modeling for Understanding Hierarchical Visual Dynamics	ICML 2026	ICML; schema-guided world modeling for hierarchical visual dynamics in multimodal LLMs
2026-07-05	SeeMe: Mitigating Hallucinations in Large Vision-Language Models through Effective Visual Token Engineering	—	Visual token engineering to mitigate hallucinations in LVLMs — practical LVLM reliability fix
2026-07-05	RL Forgets! Towards Continual Policy Optimization	—	Exposes catastrophic forgetting in RL-based VLM post-training — critical for continual adaptation
2026-07-05	ResearchStudio-Reel: Automate the Last Mile of Research from Paper to Poster, Video, and Blog	—	—
2026-07-05	AI Wizards at EXIST 2026: Hierarchical Soft-Label Learning for Multimodal Sexism Identification in Memes	—	—
2026-07-04	Do Medical Vision Language Models Actually See? A Counterfactual Grounding Framework and Hard-Negative Contrastive Training for Visually-Reliant Medical VLMs	—	Counterfactual grounding + hard-negative contrastive training exposes visual shortcut problem in medical VLMs
2026-07-04	USE: A Unified Self-Ensembling Framework for Test-Time Prompt Tuning	ICML 2026	ICML; unified self-ensembling for test-time prompt tuning advances CLIP-based TTA
2026-07-04	Bridging Interleaved Multi-Modal Reasoning as a Unified Decision Process	—	RL over interleaved text-image generation; directly informs training unified multimodal agents
2026-07-04	Look Before You Leap: Distilling Tree Search into Action Evaluation for Frozen VLA Models	—	Distills tree search into frozen VLA action evaluation; novel adaptation without fine-tuning
2026-07-03	RADIO1D: Elastic Representations for Condensed Vision Modeling	ICML 2026	ICML; challenges fixed 2D patch assumption in VLMs with elastic 1D representations
2026-07-03	Present but Not Remembered: Auditing How Frozen VLAs Encode, Deploy, and Steer Visual History	—	Mechanistic audit of how frozen VLAs encode visual history — informs VLA memory architecture
2026-07-03	MentalThink: Shaping Thoughts in Mental SVG World	—	Novel think-with-SVG paradigm enabling executable visual reasoning in MLLMs
2026-07-03	Layer-wise Cross-Lingual Depression Detection from Speech: Analysis with Contrastive Alignment	—	—
2026-07-02	Show Me Examples: Inferring Visual Concepts from Image Sets	—	Apple Research (Susskind/Bautista); fills fundamental VLM gap: concept inference from image sets without text
2026-07-02	Visually Grounded Self-Reflection for Vision-Language Models via Reinforcement Learning	—	Durrett lab; RL trains visual self-reflection/CoT correction — key for VLM agent reasoning chains
2026-07-02	LASER: A Corrective Lens for LVLMs via Visual Attention Preservation and Sink Suppression	ECCV 2026	ECCV 2026; LASER attention-sink suppression fixes visual forgetting in long LVLM decoding
2026-07-02	Towards Robustness against Typographic Attack with Training-free Concept Localization	ECCV 2026	ECCV 2026; training-free CLIP typographic attack defense — critical for all CLIP-encoder VLMs
2026-07-02	Learning to Move Before Learning to Do: Task-Agnostic pretraining for VLAs	—	Task-agnostic VLA pretraining; novel approach to VLA data bottleneck via motion pre-training
2026-07-02	VLAFlow: A Unified Training Framework for Vision-Language-Action Models via Co-training and Future Latent Alignment	—	VLAFlow unified VLA co-training + future latent alignment; clean ablation of pretraining paradigms
2026-07-02	MMBench-Live: A Continuously Evolving Benchmark for Multimodal Models	—	MMBench-Live; continuously evolving benchmark combats contamination — raises evaluation rigor bar
2026-07-02	Combating Textual Noise and Redundancy: Entropy-Aware Dense Visual Token Pruning	ECCV 2026	ECCV 2026; entropy-aware visual token pruning under dense/fine-grained instructions — practical efficiency
2026-07-02	Multimodal Knowledge Edit-Scoped Generalization for Online Recursive MLLM Editing	—	Online recursive MLLM editing with bounded overhead — essential for keeping deployed VLMs current
2026-07-02	Towards Real-World Ultrasound Understanding: Large Vision-Language Models from Multi-Image Examinations with Long-Form Reports	—	Multi-image ultrasound VLM with long-form reports; practical medical LVLM with realistic data regime
2026-07-02	AnyGroundBench: A Specialized-Domain Benchmark for Video Grounding in Vision-Language Models	—	—
2026-07-02	Teaching Vision-Language-Action Models What to See and Where to Look	ECCV 2026	VLA training data design for autonomous driving; addresses vision-text imbalance in VLA supervision
2026-07-02	SpaceEra++: A Unified Framework Towards 3D Spatial Reasoning in Video	—	Unified 3D spatial reasoning in video; directly relevant to embodied VLM spatial understanding
2026-07-02	DeepGaze3.5-VL: Modeling Scanpaths via Autoregressive Token Prediction	—	Autoregressive scanpath modeling via VLM token prediction; novel attention modeling approach
2026-07-02	COMFYCLAW: Self-Evolving Skill Harnesses for Image Generation Workflows	—	Self-evolving skill harnesses for image workflows; directly relevant to agent memory and reusable skills
2026-07-02	MolSight: A Graph-Aware Vision-Language Model for Unified Chemical Image Understanding	—	Graph-aware VLM for molecular images; novel architecture bridging chemistry and multimodal understanding
2026-07-02	MedStreamBench: A Time-Aware Benchmark for Streaming and Proactive Medical Video Understanding	—	Time-aware medical video benchmark; proactive prediction timing critical for clinical AI systems
2026-07-02	Seek to Segment: Active Perception for Panoramic Referring Segmentation	ECCV 2026	Active perception for panoramic segmentation; bridges VLMs and embodied AI agent perception
2026-07-02	Assessing VLM Reliability for Medical Image Quality Evaluation Under Corruption and Bias	—	VLM reliability under corruption for medical imaging; rigorous evaluation of deployment robustness
2026-07-02	SAB-LVLM: Significance-Aware Binarization for Large Vision-Language Models	—	Binarization compression for large VLMs; practical inference efficiency for deployment
2026-07-02	ProCal: Inference-Time Proposal Calibration for Open-Vocabulary Object Detection	—	Inference-time calibration for open-vocab detection; plug-in improvement without retraining
2026-07-02	FlowCIR: Semantic Transport via Flow Matching for Zero-Shot Composed Image Retrieval	—	Zero-shot composed image retrieval via flow matching; novel semantic transport approach for VLMs
2026-07-02	Rank-Then-Act: Reward-Free Control from Frame-Order Progress	—	VLM as reward-free ordinal progress scorer for control; directly relevant to agent harness design
2026-07-02	The Moving Eye: Enhancing VLA Spatial Generalization via Hybrid Dynamic Data Collection	—	Hybrid dynamic data collection for VLA spatial generalization; actionable recipe for robot VLM training
2026-07-02	LongEgoRefer: A Benchmark for Long-Form Egocentric Video Referring Expression Comprehension	ECCV 2026	ECCV egocentric video REC benchmark; tests long-form temporal grounding in VLMs
2026-07-02	Domain Generalization via Text-Anchored Information Bottleneck	ECCV 2026	Text-anchored information bottleneck for domain generalization; VLM-grounded invariant representation method
2026-07-02	LIME: Learning Intent-aware Camera Motion from Egocentric Video	—	Active vision for robotics; intent-aware camera motion from egocentric video, novel VLM application
2026-07-02	Embodied.cpp: A Portable Inference Runtime of Embodied AI Models on Heterogeneous Robots	—	Portable VLA/WAM inference runtime; directly addresses real-world embodied AI deployment gap
2026-07-02	Multimodal Fusion for Fine-Grained Classification of Breast Fibroadenoma and Phyllodes Tumors	—	Medical VLM fusion for breast tumor classification; fine-grained multimodal clinical decision support
2026-07-02	Search-based Testing of Vision Language Models for In-Car Scene Understanding	—	Search-based adversarial testing of VLMs for automotive safety; evaluation methodology contribution
2026-07-02	Efficient PEFT Methods with Adaptive Checkpointing for Vision Models and VLMs on Resource Constrained Consumer-GPUs	—	PEFT comparison for VLMs on consumer GPUs; practical fine-tuning on constrained hardware
2026-07-02	SINA: A Fully Automated Circuit Schematic Image to Netlist Generator Using Artificial Intelligence	—	VLM applied to circuit schematic-to-netlist; novel EDA domain grounding task with structured output
2026-07-02	EduArt: An educational-level benchmark for evaluating art history knowledge in large language models	—	Domain-specific benchmark for art history in LLMs; benchmark methodology over synthetic baselines
2026-07-02	Evaluating Vision-Language Models as a Zero-Shot Learning Alternative to You Only Look Once and Optical Character Recognition for Nigerian License Plate Recognition	—	Zero-shot VLM vs. YOLO+OCR for license plate recognition; narrow but concrete evaluation comparison
2026-07-02	RTE-FM-Dehazer: Radiative Transfer Equation Inspired Flow Matching for Real-World Image Dehazing	—	—
2026-07-02	VLA-Corrector: Lightweight Detect-and-Correct Inference for Adaptive Action Horizon	—	—
2026-07-01	Information-Regularized Attention for Visual-Centric Reasoning	ECCV 2026	ECCV; tackles hallucination, grounding, forgetting via attention regularization — core VLM reliability
2026-07-01	AdaBoosting Text Prompts for Vision-Language Models	ECCV 2026	ECCV 2026; AdaBoosting text prompts improves zero-shot VLM classification, practical and reproducible
2026-07-01	LeVLJEPA: End-to-End Vision-Language Pretraining Without Negatives	—	Non-contrastive VL pretraining without negatives — architectural shift away from CLIP-style objectives
2026-07-01	DART: Difficulty-Adaptive Routing for Zero-Shot Video Temporal Grounding	ECCV	ECCV; difficulty-adaptive routing for zero-shot video grounding, strong venue, novel routing idea
2026-07-01	Perceive-to-Reason: Decoupling Perception and Reasoning for Fine-Grained Visual Reasoning	—	Decouples perception/reasoning for fine-grained VQA — addresses critical VLM failure mode
2026-07-01	Domain Arithmetic: One-Shot VLA Adaptation under Environmental Shifts	ECCV 2026	ECCV 2026; domain arithmetic for VLA one-shot adaptation, directly relevant to robot VLM deployment
2026-07-01	StochasT: Learning with Stochastic Turn Depth for Visual Instruction Tuning	ECCV 2026	ECCV 2026; stochastic turn depth for visual instruction tuning addresses training-inference discrepancy
2026-07-01	MindEdit-Bench: Benchmarking Object-Level Counterfactual Spatial Reasoning in VLMs from In-the-Wild Photos	—	Counterfactual spatial reasoning benchmark probes genuine VLM scene understanding beyond observation
2026-07-01	GenAU: Language-Grounded Industrial Anomaly Understanding with Vision-Language Models	—	VLM for industrial anomaly detection+localization+explanation — practical grounded multimodal system
2026-07-01	Personalized Object Identification and Localization via In-Context Inference with Vision-Language Models	—	In-context inference for personalized object localization — few-shot VLM grounding without finetuning
2026-07-01	Selective Test-Time Debiasing for CLIP via Reward Gating	—	CLIP debiasing via reward gating; test-time, training-free; addresses VLM fairness at deployment
2026-07-01	Wake up for Touch! Mask-isolated Tactile Alignment Learning in MLLMs	ECCV 2026	Tactile modality in MLLMs; mask-isolated alignment; ECCV 2026; fills physical-grounding gap
2026-07-01	What's Hidden Matters: Identifying Planning-Critical Occluded Agents using Vision-Language Models	—	VLMs for occlusion-aware autonomous planning; novel selective attention over hidden agents
2026-07-01	DroneFINE: Domain-Aware Parameter-Efficient Fine-Tuning of Vision-Language Detectors for Drone Images	—	Domain-aware PEFT for VLM detectors on UAV imagery; parameter-efficient adaptation recipe
2026-07-01	LongVQUBench: Benchmarking Long-Term Video Quality Understanding of Vision-Language Models	—	Long-video quality understanding benchmark for LVLMs; temporal distortion evaluation gap
2026-07-01	Foundation Model-driven Key Anatomy Frame Selection for Blind-sweep Ultrasound Fetal Birth Weight Estimation	MICCAI 2026	MICCAI 2026; foundation model for ultrasound frame selection; low-resource medical VLM
2026-07-01	Human-Centric Transferable Tactile Pre-Training for Dexterous Robotic Manipulation	—	Tactile pre-training transferable to dexterous manipulation; cross-modal grounding for robotics
2026-07-01	Attribute-Prompted Kernel Hashing for Unsupervised Data-Efficient Cross-Modal Retrieval	—	Unsupervised cross-modal hashing with attribute prompts; data-efficient retrieval baseline
2026-07-01	DroneIQA-VLE: Multi-Task Drone Image Quality Assessment via Vision-Language Ensemble	—	VLM ensemble for drone image quality assessment; ICME 2026 challenge solution
2026-07-01	Discrete Diffusion Language Models for Interactive Radiology Report Drafting	—	—
2026-07-01	Disentangling Pictorial Cue Understanding from Language Bias in VLMs via Depth Ordering Task	ECCV 2026	Probes VLM vision vs. language bias on depth perception; diagnostic for building reliable VLMs
2026-07-01	CPG-PAD: Concept-Informed Prompts Guided Presentation Attack Detection	—	Concept-informed prompting for face PAD; practical VLM prompting technique for security-critical vision
2026-07-01	ESC: Emotional Self-Correction for Reliable Vision-Language Models	ECCV	Emotional self-correction for VLMs without post-training; lightweight reliability fix applicable broadly
2026-07-01	GMO-E\(^2\)DIT: Grounded Multi-Operation Editing for E-Commerce Images	—	Multi-operation grounded image editing for e-commerce; tests compositional VLM instruction following
2026-07-01	MIBE: Multi-subject Interaction Benchmark and Evaluator for Personalized Image Generation	—	Multi-subject personalized image generation benchmark; evaluates identity-binding in generative VLMs
2026-07-01	Multi-modal Rail Crossing Safety Analysis	—	Multimodal safety analysis fusing vision and structured accident reports; applied safety-critical domain
2026-07-01	Towards Developing a Multimodal Chat Assistant for University Stakeholders: RAG-based Approach	—	RAG-based multimodal university assistant; practical retrieval-augmented VLM deployment case study

June 2026 (109)

Date	Paper	Venue	Why selected
2026-06-30	One Video, One World: Turning Monocular Video into Physical 4D Scenes	ECCV 2026	Saining Zhang; first training-free 4D mesh scene from single video; highly novel
2026-06-30	Decodable Is Not Grounded: A Vision-Ablation Arbiter for VLM Spatial Reasoning	—	Tat-Seng Chua; probes overestimate VLM grounding — critical eval result
2026-06-30	Harnessing Textual Refusal Directions for Multimodal Safety	—	Nicu Sebe; refusal-direction safety for MLLMs without unsafe data; timely
2026-06-30	Attend, Transform, or Silence: Operator-Level Visual Skipping for Efficient Multimodal LLM Inference	—	Operator-level visual token skipping; directly cuts MLLM inference cost
2026-06-30	Evidence Triangulation for Multimodal Fact-Checking in the Wild	—	Papadopoulos; multimodal fact-checking in the wild; real-world relevance
2026-06-30	HyperVLP: Enhancing Hierarchical Surgical Video-Language Pre-training in Hyperbolic Space	—	Nassir Navab; surgical VLM pretraining in hyperbolic space; strong med contribution
2026-06-30	Language-Assisted Super-Resolution from Real-World Low-Resolution Patches	—	Kyoung Mu Lee; language-guided real-world SR; creative cross-modal approach
2026-06-30	RCT: A Robot-Collected Touch-Vision-Language Dataset for Tactile Generalization	—	Calandra; robot-collected tactile-vision-language dataset; enables tactile VLMs
2026-06-30	Breaking Failure Cascades: Step-Aware Reinforcement Learning for Medical Multimodal Reasoning	—	Step-aware RL for medical VLM reasoning; fixes failure cascades in multimodal CoT
2026-06-30	Token-Sparse Medical Multimodal Reasoning via Dual-Stream Reinforcement Learning	—	Dual-stream RL for token-sparse medical VLM reasoning; novel sparse-evidence approach
2026-06-30	Visual Semantic Entropy: Do Vision Language Models Recognize Visual Ambiguity?	—	Visual ambiguity detection for VLMs; essential for reliable deployed systems
2026-06-30	Technical Report of RoboSpatial Challenge at CVPR 2026: Selective Reasoning Activation and Reference-Frame Disambiguation for Embodied Spatial Reasoning	—	CVPR 2026; embodied spatial reasoning benchmark + winning solution
2026-06-30	Xiaomi-GUI-0 Technical Report	—	Industry-scale GUI agent from Xiaomi; directly applicable to agent building
2026-06-30	Reasoning-aware Speculative Decoding for Efficient Vision-Language-Action Models in Autonomous Driving	—	Speculative decoding for VLA reasoning; practical latency reduction technique
2026-06-30	Think While You Map: Asynchronous Vision-Language Agents for Incremental 3D Scene Graphs	ECCV 2026	ECCV 2026; asynchronous incremental 3D scene graphs for embodied agents
2026-06-30	Rethinking Foundation Model Collaboration: Enhancing Specialized Models through Proxy Task Reasoning	—	FM + specialized model collaboration via proxy tasks; key agent architecture pattern
2026-06-30	Seeing Is Not Sharing: Some Vision-Language Models Overestimate Common Ground in Asymmetric Dialogue	—	—
2026-06-30	InstanceControl: Controllable Complex Image Generation without Instance Labeling	—	—
2026-06-30	3D HAMSTER: Bridging Planning and Control in Hierarchical Vision Language Action Models through 3D Trajectory Guidance	—	—
2026-06-29	H-GRPO: Permutation-Invariant Reinforcement Learning for Grounded Visual Reasoning	—	GRPO + grounded VLM reasoning; directly relevant to RL-based VLM improvement
2026-06-29	VisReflect: Latent Visual Reflection for Fine-Grained Perception in Long Visual Context	ECCV 2026	Addresses the long-visual-context bottleneck that limits real-world VLM deployment
2026-06-29	MemLeak: Diagnosing Information Leaks in Multimodal Agent Memory	—	Surprising finding on recoverable visual data after memory deletion; security implications
2026-06-29	See Only When Needed: Context-Aware Attention Intervention for Mitigating Hallucinations in LVLMs	ECCV 2026	Snoek group; ECCV 2026; training-free hallucination mitigation without amplifying spurious regions
2026-06-29	Training Vision-Language-Action Models with Dense Embodied Chain-of-Thought Supervision	—	Dense embodied CoT supervision tackles the cross-embodiment bottleneck in VLA models
2026-06-29	From Accuracy to Visual Dependence: Auditing and Filtering Modality Collapse in Traffic VideoQA	—	Akata group; fundamental audit of whether VLMs truly use visual evidence
2026-06-29	Same Concept, Different Directions: Cross-Modal Feature Heterogeneity in Sparse Autoencoders	—	First sparse autoencoder analysis of cross-modal feature heterogeneity in VLMs
2026-06-29	GROW\(^2\): Grounding Which and Where for Robot Tool Use	—	David Hsu; open-world affordance grounding for creative robot tool use with VLMs
2026-06-29	UnfoldArt: Zero-Shot Recovery of Full Articulated 3D Objects from Text or Image	—	Laptev + Angela Dai; zero-shot articulated 3D recovery; crucial for embodied AI
2026-06-29	The Label Imitation Game: Turing Test Network for Zero-Shot Pseudo-Label Pruning	ECCV 2026	Jason Corso; Turing-test method cleans zero-shot pseudo-labels; novel + practical
2026-06-29	Automating the Design of Embodied AgentArchitectures	—	Automated embodied agent architecture design; changes how we build agent harnesses
2026-06-29	DOPD: Dual On-policy Distillation	—	—
2026-06-29	Unlocking the Visual Record of Materials Science: A Large-Scale Multimodal Dataset from Scientific Literature	—	—
2026-06-29	JD Oxygen AI Item Center (Oxygen AIIC) V1: An Industrial-Scale LLM/VLM-Centric Solution for Item Understanding, Management, and Applications	—	—
2026-06-28	ScAle: Attention Head Scaling as a Minimal Adapter for Spatial Reasoning in Vision Language Models	ECCV 2026	Minimal efficient adapter for spatial reasoning, a known VLM weakness; Yanzhi Wang group
2026-06-28	MIRROR: Aligning Semantic Relations from Language to Image via Gromov--Wasserstein	ECCV 2026	Fundamental structural alignment gap in MLLMs; novel Gromov-Wasserstein approach
2026-06-28	Fast Enough to Act: Spatio-Temporal Visual Token Merging for Low-Latency Robotic VLMs and VLAs	—	Token merging for low-latency robotic VLM inference; practical efficiency contribution
2026-06-28	Event-VLA: Action-Conditioned Event Fusion for Robust Vision-Language-Action Model	—	Event camera fusion addresses RGB degradation in VLA; novel sensor modality for embodied AI
2026-06-28	Rank-Aware Hyperbolic Alignment for Vision-Language Dataset Distillation	—	—
2026-06-27	On Test-Time Scaling for Vision-Language Models	ECCV 2026	Test-time scaling is the hottest topic; timely application to VLMs is high-impact
2026-06-27	HKVLM: Faithful Reasoning Grounding by Binding Language Queries to a Frozen Detector	—	Practical binding of LM reasoning to frozen detector; addresses VLM faithfulness directly
2026-06-27	ViPSim: Collaborating Visual and Parameter Spaces for Consistent Long-Horizon Embodied World Models	—	Consistent world models for evaluating VLA systems; critical embodied AI infrastructure
2026-06-26	Vision-Default, Prior-Override: Causal Mechanisms of Perception-Knowledge Conflict in Vision-Language Models	—	Mechanistic causal account of perception-knowledge conflict; VLM reliability foundation
2026-06-26	AirGroundBench: Probing Spatial Intelligence in Multimodal Large Models under Heterogeneous Multi-View Embodied Collaboration	—	ECCV; spatial reasoning benchmark for MLLMs across heterogeneous multi-view embodied settings
2026-06-26	HAT-4D: Lifting Monocular Video for 4D Multi-Object Interactions via Human-Agent Collaboration	ECCV 2026	ECCV; monocular→4D pipeline for scalable VLA training data at low cost
2026-06-26	MVPruner: Dynamic Token Pruning for Accelerating Multi-view Vision-Language Models in Autonomous Driving	—	Dynamic token pruning for multi-view VLMs; practical efficiency for autonomous driving
2026-06-26	RSICCLLM: A Multimodal Large Language Model for Remote Sensing Image Change Captioning	ECCV 2026	—
2026-06-26	EchoSonar-R: A Multi-View Reasoning-Enabled Model for Disease Classification and Report Generation in Echocardiography	—	—
2026-06-26	Translation as a Bridging Action: Transferring Manipulation Skills from Humans to Robots	—	—
2026-06-26	ReScene: Structured Indoor Scene Reconstruction from Multi-View Captures	—	—
2026-06-26	SpatialUAV: Benchmarking Spatial Intelligence for Low-Altitude UAV Perception, Collaboration, and Motion	—	—
2026-06-26	Panoramic Scene Analysis: A Survey from Distortion-Aware Engineering to Sphere-Native Foundation Modeling	—	—
2026-06-26	Enhancing Numerical Prediction in LLMs via Smooth MMD Alignment	—	—
2026-06-26	ZooClaw-FashionSigLIP2: Distilled Fine-tuning for Robust Fashion Retrieval	—	—
2026-06-26	VLM-Aware Meta-Optic Front-End Design for Frozen Vision-Language Models	—	—
2026-06-26	DataComp-VLM: Improved Open Datasets for Vision-Language Models	—	DataComp series has proven impact; open VLM dataset benchmark is foundational for the whole field
2026-06-26	Animation2Code: Evaluating Temporal Visual Reasoning in Video-to-Code Generation	—	Temporal visual reasoning evaluation gap; Alane Suhr; tests if VLMs understand dynamics
2026-06-26	Detecting Clinical Hallucinations in LVLMs via Counterfactual Visual Grounding Uncertainty	—	Counterfactual visual grounding for clinical hallucination detection; important safety work
2026-06-26	Drop-Then-Recovery: How Redundant Are Vision-Language-Action Models?	—	Systematic VLA redundancy analysis; actionable pruning for efficient deployment
2026-06-25	From Hallucination to Grounding: Diagnosing Visual Spatial Intelligence via CRISP	ECCV 2026	ECCV 2026; structural diagnostic separating language priors from true spatial reasoning in VLMs
2026-06-25	GAVEL: Grounded Caption Error Verification and Localization	—	VLM hallucination detection+localization; directly useful for grounded VLM eval pipelines
2026-06-25	Paying More Attention to Visual Tokens in Self-Evolving Large Multimodal Models	ECCV 2026	ECCV 2026; self-evolving LMM training with visual token attention — unsupervised reasoning improvement
2026-06-25	Do Image Editing Models Understand Lighting?	—	Nießner+Rother at ECCV; probes fundamental lighting understanding gap in generative VLMs
2026-06-25	Improving Vision-Language-Action Model Fine-Tuning with Structured Stage and Keyframe Supervision	—	Structured stage+keyframe supervision for VLA fine-tuning; practical reproducible recipe
2026-06-25	Bridging Vision and Language Concepts through Optimal Transport Semantic Flow	—	OT-based visual-textual concept alignment; directly improves interpretable VLM reasoning in CBMs
2026-06-25	HarmVideoBench: Benchmarking Harmful Video Understanding in Large Multimodal Models	—	Systematic harmful-video benchmark for LMMs; fills critical safety evaluation gap for video VLMs
2026-06-25	Adversarial Diffusion Across Modalities: A Fusion Survey of Attacks, Defenses, and Evaluation for Text, Vision, and Vision-Language Models	—	Comprehensive survey unifying adversarial attacks across text/vision/VLMs; useful robustness reference
2026-06-25	Dismantling Pathological Shortcuts: A Causal Framework for Faithful LVLM Decoding	ICML 2026	ICML; causal root-cause of LVLM hallucination beyond attention heuristics
2026-06-25	Large Language Model Teaches Visual Students: Cross-Modality Transfer of Fine-Grained Conceptual Knowledge	ICML 2026	Yong Jae Lee, ICML; LLM→vision cross-modal fine-grained knowledge distillation
2026-06-25	Aloe-Vision: Robust Vision-Language Models for Healthcare	MIDL 2026	MIDL; robustness-focused healthcare LVLM with scarce-data evaluation
2026-06-24	Disease-Centric Vision-Language Pretraining with Hybrid Visual Encoding for 3D Computed Tomography	ICML 2026	ICML; 3D CT VLP with hybrid encoding; directly buildable medical foundation recipe
2026-06-24	MRI2Rep: Autoregressive Structured Report Generation for 3D Liver MRI	MICCAI 2026	MICCAI; Zongwei Zhou; autoregressive structured MRI reports end-to-end
2026-06-24	Beyond Visual Forensics: Auditing Multimodal Robustness for Synthetic Medical Image Detection	MICCAI 2026	MICCAI; Pin-Yu Chen; audits VLM robustness on synthetic medical images
2026-06-24	Physics Question Scene Graph: Fine-grained Evaluation of Physical Plausibility in Text-to-Video Generation	ECCV 2026	Bansal group (UNC, ECCV 2026); scene-graph eval for physical plausibility in T2V — new rigorous eval tool
2026-06-24	H-Adapter: Pose-Robust Hairstyle Transfer via Attention-Derived, Source-Aligned Hair Masks	ECCV 2026	ECCV 2026; attention-derived masks for pose-robust image editing with VLM-adjacent mechanisms
2026-06-24	Invoice Haystack: Benchmarking Document Retrieval and Visual Question Answering Under Strong Visual Homogeneity	ECCV 2026	ECCV; benchmark exposing VLM retrieval failures under visual homogeneity at scale
2026-06-23	ActiveScope: Actively Seeking and Correcting Perception for MLLMs	ICML 2026	ICML 2026; training-free fine-grained perception correction for MLLMs — drop-in applicable
2026-06-23	VisCritic: Visual State Comparison as Process Reward for GUI Agents	ECCV 2026	ECCV 2026; visual-state process reward for GUI agents — directly useful for agent eval harnesses
2026-06-23	BioMedVR: Confusion-Aware Mixture-of-Prompt Experts for Biomedical Visual Reprogramming	ECCV 2026	ECCV 2026; confusion-aware mixture-of-prompt experts for biomedical CLIP adaptation
2026-06-23	EgoSAT: A Comprehensive Benchmark of Egocentric Streaming Interaction Understanding	ECCV 2026	Yin Li group (ECCV 2026); first streaming egocentric VLM benchmark — critical for real-time agent evaluation
2026-06-23	Evaluating the Interpretability of Sparse Autoencoders with Concept Annotations	ECCV 2026	ECCV 2026; first concept-annotation-grounded evaluation of SAE interpretability in VLMs
2026-06-23	Transition-Aware best-of-N sampling for Longitudinal Chest X-ray Reports	—	—
2026-06-22	CFPO: Counterfactual Policy Optimization for Multimodal Reasoning	ICML 2026	ICML 2026; counterfactual causal RL for VLM reasoning—novel mechanism beyond standard GRPO/PPO
2026-06-22	Dense Reward for Multi-View 3D Reasoning with Global Maps and Local Views	ECCV 2026	ECCV 2026; dense reward shaping for multi-view 3D VQA—novel RL signal for spatial chain-of-thought
2026-06-22	Brain-Adapter: A Dual-Stream Vision-Language MIL Framework for Comprehensive 3D CT Diagnosis of Acute Intracranial Pathologies	MICCAI 2026	MICCAI 2026; dual-stream MIL VLM for 3D CT; directly applicable to medical foundation model work
2026-06-22	Mind the Heads: Topological Representation Alignment for Multimodal LLMs	—	—
2026-06-21	PolicyTrim: Boosting Intrinsic Policy Efficiency of Vision-Language-Action Models	ECCV 2026	ECCV 2026; intrinsic policy efficiency for VLA deployment; directly buildable
2026-06-18	Timage: A Generative Text-in-Image Paradigm for Fine-Tuning Vision-Language Models	ECCV	ECCV; text-in-image geometric anchoring directly addresses fine-grained spatial reasoning gap in MLLMs
2026-06-18	Confidence Calibration for Multimodal LLMs: An Empirical Study through Medical VQA	MICCAI 2025	MICCAI 2025; confidence calibration mismatch in medical MLLMs creates direct misdiagnosis risk
2026-06-18	FOCA: Future-Oriented Conditioning for Data-Efficient Vision-Language-Action Adaptation	ICML 2026	ICML 2026; systematic few-shot stress test of VLA adaptation with future-oriented conditioning
2026-06-18	Scalable Training of Spatially Grounded 2D Vision-Language Models for Radiology	MICCAI 2026	MICCAI 2026; 1.2M CT/MR grounded pairs; scalable spatially grounded radiology VLM recipe
2026-06-17	Moving Beyond Diversity: Visual Token Pruning as Subspace Reconstruction for Efficient VLMs	ECCV 2026	ECCV 2026; novel subspace-reconstruction reframing of token pruning; critical efficiency gain
2026-06-17	Semantic Robustness Certification for Vision-Language Models	ICML	ICML; first certified robustness framework for semantic distribution shifts in VLMs
2026-06-17	Hallucination Detection and Correction in Medical VLMs via Counter-Evidence Verification	MICCAI 2026	MICCAI 2026; counter-evidence verification for hallucinations in high-stakes medical VLMs
2026-06-17	Bridging Creative Intent and Visual Quality: Creator-Driven Recurrent Video Generation with Agentic Feedback Loops	ICML 2026	ICML 2026 acceptance; agentic feedback loop for coherent long-form video generation
2026-06-17	Language-Instructed Vision Embeddings for Controllable and Generalizable Perception	ICLR 2026	ICLR 2026; novel paradigm using language to steer vision embeddings for controllable perception
2026-06-16	Visuals Lie, Consistency Speaks: Disentangling Spatial Attention from Reliability in Vision-Language Models	ICLR 2026	ICLR 2026; challenges Attention-Confidence Assumption; exposes core VLM reliability blind spot
2026-06-16	MODE-RAG: Manifold Outlier Diagnosis and Energy-based Retrieval-Augmented Generation Evaluation	ACL 2026	ACL 2026; multimodal RAG evaluation exposing cross-modal hallucination and sycophancy failures
2026-06-16	Evaluating Large Language Models Abilities for Addressee, Turn-change, and Next Speaker Prediction in Meetings	INTERSPEECH 2026	—
2026-06-16	Domain Generalizable Adaptation of 3D Vision-Language Models via Regularized Fine-Tuning	TMLR	TMLR; regularized fine-tuning for domain-generalizable 3D vision-language model adaptation
2026-06-15	Lost at the End: Primacy Bias in Multimodal Retrieval-Augmented Question Answering	EMNLP 2026	—
2026-06-15	XAI-Grounded Explanation Generation for Speech Deepfake Detection with Training-Free Multimodal Large Language Models	INTERSPEECH 2026	—
2026-06-14	Learning Directional Semantic Transitions for Longitudinal Chest X-ray Analysis	MICCAI 2026	—
2026-06-14	The Truth Stays in the Family: Enhancing Contextual Grounding via Inherited Truthful Heads in Model Lineages	ICML 2026	ICML 2026; inherited truthful heads across model lineages offers new grounding tool against hallucination
2026-06-13	Keep It in Mind: User Centric Continual Spatial Intelligence Reasoning in Egocentric Video Streams	ICML	—
2026-06-13	Think Less, Act Early: Reinforced Latent Reasoning with Early Exit in Vision-Language-Action Models	ICML 2026	ICML 2026; reinforced latent reasoning with early exit cuts VLA compute and error propagation
2026-06-12	CORA: Analyzing and bridging thinking-answer gap in Multimodal RLVR via Consistency-Oriented Reasoning Alignment	EMNLP 2026	—
2026-06-12	One Layer's Trash is Another Layer's Treasure: Adaptive Layer-wise Visual Token Selection in LVLMs	CVPR 2026	—
2026-06-11	Cross-Modal Masked Compositional Concept Modeling for Enhancing Visio-Linguistic Compositionality	ACL 2026	—
2026-06-11	OpenMedQ: Broad Open Pretraining for Medical Vision-Language Models	MIDL	—

May 2026 (20)

Date	Paper	Venue	Why selected
2026-05-30	Decomposed On-Policy Distillation for Vision-Language Reasoning: Steering Gradients for Visual Grounding	ICML 2026	ICML 2026; decomposes on-policy distillation gradients for VL visual grounding
2026-05-30	Through the PRISM: Principle-Aware, Interpretable, and Multi-Scale Evaluation of Visual Designs	CVPR 2026	CVPR 2026; PRISM: principle-aware multi-scale evaluation of visual designs
2026-05-29	Learning from Fine-Grained Visual Discrepancies: Mitigating Multimodal Hallucinations via In-Context Visual Contrastive Optimization	ICML 2026	ICML 2026; visual contrastive DPO for multimodal hallucination mitigation
2026-05-29	Diversity Matters: Revisiting Test-Time Compute in Vision-Language Models	ICML 2026	Mrinmaya Sachan (ETH); ICML 2026; systematic study of test-time compute in VLMs
2026-05-29	How can embedding models bind concepts?	ICML 2026	Seong Joon Oh (Tübingen); ICML 2026; concept binding failure in CLIP-style models
2026-05-29	Immuno-VLM: Immunizing Large Vision-Language Models via Generative Semantic Antibodies for Open-World Trustworthiness	ICML 2026	ICML 2026; immunizing VLMs against open-world adversarial semantic vulnerabilities
2026-05-28	Grounded 3D-Aware Spatial Vision-Language Modeling	CVPR 2026	Ligeng Zhu (MIT/NVIDIA); CVPR 2026; unified 2D+3D grounding in spatial VLMs
2026-05-28	Beyond 3D VQAs: Injecting 3D Spatial Priors into Vision-Language Models for Enhanced Geometric Reasoning	CVPR 2026	Yi Ma (Berkeley) + Joseph Tighe (Meta); CVPR 2026; 3D spatial priors improve VLM geometry
2026-05-28	Unveiling the Visual Counting Bottleneck in Vision-Language Models	ICML 2026	Mrinmaya Sachan (ETH); ICML 2026; reveals visual counting extrapolation bottleneck
2026-05-28	Improving CLIP Adaptation by Breaking Tail Alignment for Source-Free Cross-Domain Few-Shot Learning	ICML 2026	ICML 2026; breaks tail alignment in CLIP for cross-domain few-shot learning
2026-05-27	Self-Prophetic Decoding to Unlock Visual Search in LVLMs	ICML 2026	ICML 2026; self-prophetic decoding enables visual search in LVLMs
2026-05-24	Interpretability Transfer from Language to Vision via Sparse Autoencoders	ICML 2026	ICML 2026; sparse autoencoders transfer language interpretability to vision
2026-05-24	Language Bias in LVLMs: From In-Depth Analysis to Simple and Effective Mitigation	ICML 2026	ICML 2026; in-depth analysis + simple mitigation of language bias in LVLMs
2026-05-22	CVSearch: Empowering Multimodal LLMs with Cognitive Visual Search for High-Resolution Image Perception	ICML 2026	ICML 2026; cognitive visual search for high-resolution image perception in MLLMs
2026-05-22	Multimodal Distribution Matching for Vision-Language Dataset Distillation	CVPR 2026	CVPR 2026; multimodal distribution matching for vision-language dataset distillation
2026-05-22	Turning Adaptation into Assets: Cross-Domain Bridging for Online Vision-Language Navigation	ICML 2026	ICML 2026; test-time adaptation for VLN under non-stationary distribution shifts
2026-05-22	SPACENUM: Revisiting Spatial Numerical Understanding in VLMs	—	—
2026-05-21	Seeing without Looking: Do Vision-Language Benchmarks Really Test Vision?	CVPR 2026	CVPR 2026; reveals VLM benchmark scores don't require grounded visual evidence
2026-05-21	From Abstraction to Instantiation: Learning Behavioral Representation for Vision-Language-Action Model	ICML 2026	ICML 2026; learning generalized behavioral representations for VLA distribution shift
2026-05-21	AwareVLN: Reasoning with Self-awareness for Vision-Language Navigation	CVPR 2026	CVPR 2026; self-awareness in VLN agents for reasoning over visual environments

April 2026 (33)

Date	Paper	Venue	Why selected
2026-04-30	Improving Calibration in Test-Time Prompt Tuning for Vision-Language Models via Data-Free Flatness-Aware Prompt Pretraining	CVPR 2026	Flatness-aware pretraining improves calibration in VLM test-time prompt tuning
2026-04-29	Beyond Shortcuts: Mitigating Visual Illusions in Frozen VLMs via Qualitative Reasoning	CVPR	Qualitative reasoning mitigates optical illusion failures in frozen VLMs; CVPR
2026-04-28	Prefill-Time Intervention for Mitigating Hallucination in Large Vision-Language Models	CVPR 2026	Novel prefill-time intervention eliminates LVLM hallucinations; inference-time method
2026-04-27	Improving Vision-language Models with Perception-centric Process Reward Models	CVPR	Process-level reward models for VLM RLVR; finer than outcome-only supervision
2026-04-27	LearnPruner: Rethinking Attention-based Token Pruning in Vision Language Models	ICLR 2026	Attention-based token pruning rethought; attention scores don't indicate importance
2026-04-27	Jailbreaking Frontier Foundation Models Through Intention Deception	CVPR 2026	Jailbreaks via intention deception bypass safety training; CVPR 2026 novel vector
2026-04-27	ReVSI: Rebuilding Visual Spatial Intelligence Evaluation for Accurate Assessment of VLM 3D Reasoning	ICML 2026	Rebuilds 3D spatial reasoning evaluation exposing VLM annotation biases; ICML 2026
2026-04-27	Libra-VLA: Achieving Learning Equilibrium via Asynchronous Coarse-to-Fine Dual-System	ACL 2026	Asynchronous coarse-to-fine dual system balances VLA learning equilibrium; ACL 2026
2026-04-27	SMoES: Soft Modality-Guided Expert Specialization in MoE-VLMs	CVPR 2026	Soft modality routing in MoE-VLMs; first study of modality-specific expert allocation
2026-04-27	World-R1: Reinforcing 3D Constraints for Text-to-Video Generation	ICML 2026	RL reinforces 3D geometric consistency in text-to-video generation; ICML 2026
2026-04-25	Exploring Hierarchical Consistency and Unbiased Objectness for Open-Vocabulary Object Detection	CVPR 2026	Hierarchical consistency with unbiased objectness for CLIP open-vocab detection
2026-04-24	DocPrune:Efficient Document Question Answering via Background, Question, and Comprehension-aware Token Pruning	CVPR 2026	Background/question-aware token pruning for efficient document VQA; CVPR 2026
2026-04-23	Ramen: Robust Test-Time Adaptation of Vision-Language Models with Active Sample Selection	CVPR 2026	Active sample selection for robust CLIP test-time adaptation; CVPR 2026
2026-04-22	Building a Precise Video Language with Human-AI Oversight	CVPR 2026	CVPR 2026 scalable video-language human-AI oversight; datasets and benchmarks
2026-04-22	Mitigating Hallucinations in Large Vision-Language Models without Performance Degradation	ACL 2026	Hallucination-free fine-tuning without performance degradation; ACL 2026
2026-04-22	Breaking the Illusion: When Positive Meets Negative in Multimodal Decoding	CVPR 2026	Training-free positive-negative decoding reduces object hallucination; CVPR 2026
2026-04-22	OMIBench: Benchmarking Olympiad-Level Multi-Image Reasoning in Large Vision-Language Model	ACL 2026	Olympiad-level multi-image reasoning benchmark for frontier VLMs; ACL 2026
2026-04-22	Evian: Towards Explainable Visual Instruction-tuning Data Auditing	ACL 2026	Explainable visual instruction data auditing for LVLM training quality
2026-04-21	CAPruner: Conceptual-Adjacent Scene Graph Pruner for Enhancing 3D Spatial Reasoning of Large Language Models	ACL 2026	Scene graph pruning removes irrelevant anchors for LLM 3D spatial reasoning
2026-04-21	Environmental Understanding Vision-Language Model for Embodied Agent	CVPR	Environmental understanding VLM improves instruction-following embodied agent generalization
2026-04-20	Hierarchically Robust Zero-shot Vision-language Models	CVPR 26	Hierarchically robust VLM fine-tuning preserves zero-shot under adversarial attack
2026-04-20	S2H-DPO: Hardness-Aware Preference Optimization for Vision-Language Models	ACL 2026	Hardness-aware DPO closes multi-image reasoning gap in VLM alignment
2026-04-20	Test-Time Perturbation Learning with Delayed Feedback for Vision-Language-Action Models	CVPR 2026	Test-time perturbation learning corrects VLA trajectory overfitting; CVPR 2026
2026-04-20	Enhancing Continual Learning of Vision-Language Models via Dynamic Prefix Weighting	CVPR 2026	Dynamic prefix weighting for domain-class incremental VLM learning; CVPR 2026
2026-04-20	From Heads to Neurons: Causal Attribution and Steering in Multi-Task Vision-Language Models	ACL 2026	Neuron-level causal attribution reveals cross-task VLM mechanisms; ACL 2026
2026-04-19	More Than Meets the Eye: Measuring the Semiotic Gap in Vision-Language Models via Semantic Anchorage	ACL 2026	High visual fidelity hinders abstract VLM reasoning; semiotic gap study
2026-04-19	Waking Up Blind: Cold-Start Optimization of Supervision-Free Agentic Trajectories for Grounded Visual Perception	ACL 2026	Cold-start agentic VLM grounding without expensive supervised trajectory tuning
2026-04-18	EvoComp: Learning Visual Token Compression for Multimodal Large Language Models via Semantic-Guided Evolutionary Labeling	CVPR 2026	Evolutionary semantic labeling for visual token compression; CVPR 2026 efficiency
2026-04-18	SIF: Semantically In-Distribution Fingerprints for Large Vision-Language Models	CVPR 2026	In-distribution semantic fingerprints for VLM IP protection; CVPR 2026
2026-04-17	Aligning What Vision-Language Models See and Perceive with Adaptive Information Flow	CVPR 2026	Adaptive info flow fixes VLMs that see but don't perceive; CVPR 2026
2026-04-17	TTL: Test-time Textual Learning for OOD Detection with Pretrained Vision-Language Models	CVPR 2026	Test-time textual learning improves CLIP-based OOD detection without training
2026-04-17	Where Do Vision-Language Models Fail? World Scale Analysis for Image Geolocalization	CVPR	World-scale geolocalization analysis reveals systematic VLM perception failures; CVPR
2026-04-15	One Token per Highly Selective Frame: Towards Extreme Compression for Long Video Understanding	NEURIPS 2025	Extreme compression to one token per frame enables long-video VLMs; NeurIPS 2025

March 2026 (43)

Date	Paper	Venue	Why selected
2026-03-31	Robust Multimodal Safety via Conditional Decoding	ACL 2026	ACL 2026; conditional decoding restores safety alignment in multimodal LLMs
2026-03-31	A Comprehensive Information-Decomposition Analysis of Large Vision-Language Models	ICLR 2026	ICLR 2026; information-decomposition reveals if VLMs use true multimodal fusion or shortcuts
2026-03-31	Hierarchical Pre-Training of Vision Encoders with Large Language Models	CVPR	CVPR; hierarchical joint pre-training tightly couples vision encoders with LLMs
2026-03-31	AGFT: Alignment-Guided Fine-Tuning for Zero-Shot Adversarial Robustness of Vision-Language Models	CVPR 2026	CVPR 2026; alignment-guided fine-tuning preserves zero-shot adversarial robustness in VLMs
2026-03-30	Explaining CLIP Zero-shot Predictions Through Concepts	CVPR 2026	CVPR 2026; concept bottleneck explanations make CLIP zero-shot predictions transparent
2026-03-30	Learning Multi-View Spatial Reasoning from Cross-View Relations	CVPR 2026	CVPR 2026; cross-view relational learning for multi-view spatial reasoning in VLMs
2026-03-30	ManipArena: Comprehensive Real-world Evaluation of Reasoning-Oriented Generalist Robot Manipulation	CVPR 2026	CVPR 2026; comprehensive real-world benchmark for VLA manipulation evaluation
2026-03-30	FlowHijack: A Dynamics-Aware Backdoor Attack on Flow-Matching Vision-Language-Action Models	CVPR 2026	CVPR 2026; first backdoor attack targeting flow-matching VLA models like π0
2026-03-30	\(AutoDrive\text{-}P^3\): Unified Chain of Perception-Prediction-Planning Thought via Reinforcement Fine-Tuning	ICLR 2026	ICLR 2026; chain of perception-prediction-planning with RL for autonomous driving VLMs
2026-03-29	On Token's Dilemma: Dynamic MoE with Drift-Aware Token Assignment for Continual Learning of Large Vision Language Models	CVPR 2026	CVPR 2026; drift-aware dynamic MoE token assignment for continual multimodal learning
2026-03-28	Understanding and Mitigating Hallucinations in Multimodal Chain-of-Thought Models	CVPR 2026	CVPR 2026; diagnoses root cause and mitigates hallucination in multimodal chain-of-thought
2026-03-28	Diagnosing and Repairing Unsafe Channels in Vision-Language Models via Causal Discovery and Dual-Modal Safety Subspace Projection	CVPR 2026	CVPR 2026; causal discovery identifies unsafe channels; dual-modal safety subspace repair
2026-03-28	Structural Graph Probing of Vision-Language Models	CVPR	CVPR; structural graph probing reveals neural topology organization in VLMs
2026-03-28	SpatialStack: Layered Geometry-Language Fusion for 3D VLM Spatial Reasoning	CVPR 2026	CVPR 2026; layered geometry-language fusion for reliable 3D spatial VLM reasoning
2026-03-28	ChartNet: A Million-Scale, High-Quality Multimodal Dataset for Robust Chart Understanding	CVPR 2026	CVPR 2026; million-scale high-quality chart dataset for geometry-numerical-language reasoning
2026-03-28	Seeing the Scene Matters: Revealing Forgetting in Video Understanding Models with a Scene-Aware Long-Video Benchmark	CVPR 2026	CVPR 2026; scene-aware benchmark reveals catastrophic forgetting in long-video VLMs
2026-03-28	VIRST: Video-Instructed Reasoning Assistant for SpatioTemporal Segmentation	CVPR 2026	CVPR 2026; VLM-driven spatiotemporal reasoning for referring video object segmentation
2026-03-28	Uni-World VLA: Interleaved World Modeling and Planning for Autonomous Driving	ECCV 2026	ECCV 2026; interleaved world modeling and planning unifies AD perception and action
2026-03-28	MEDIC-AD: Towards Medical Vision-Language Model's Clinical Intelligence	CVPR 2026	CVPR 2026; medical VLM with lesion detection, tracking, and visual explainability
2026-03-27	A Provable Energy-Guided Test-Time Defense Boosting Adversarial Robustness of Large Vision-Language Models	CVPR	CVPR; provable energy-guided test-time defense for VLM adversarial robustness
2026-03-27	HandVQA: Diagnosing and Improving Fine-Grained Spatial Reasoning about Hands in Vision-Language Models	CVPR 2026	CVPR 2026; benchmark diagnosing fine-grained hand spatial reasoning failures in VLMs
2026-03-27	Disentangled Robot Learning via Separate Forward and Inverse Dynamics Pretraining	ICLR 2026	ICLR 2026; disentangled forward/inverse dynamics pretraining resolves VLA misalignment
2026-03-27	Towards GUI Agents: Vision-Language Diffusion Models for GUI Grounding	CVPR 2026	CVPR 2026; diffusion VLMs for GUI grounding challenge autoregressive model dominance
2026-03-27	FairLLaVA: Fairness-Aware Parameter-Efficient Fine-Tuning for Large Vision-Language Assistants	CVPR 2026	CVPR 2026; fairness-aware PEFT for multimodal LLMs in clinical high-stakes settings
2026-03-26	No Hard Negatives Required: Concept Centric Learning Leads to Compositionality without Degrading Zero-shot Capabilities of Contrastive Models	CVPR 2026	CVPR 2026; concept-centric learning enables compositionality without zero-shot degradation
2026-03-26	HiSpatial: Taming Hierarchical 3D Spatial Understanding in Vision-Language Models	CVPR 2026	CVPR 2026; hierarchical 3D spatial understanding with perception-to-reasoning pipeline
2026-03-26	Drive My Way: Preference Alignment of Vision-Language-Action Model for Personalized Driving	CVPR 2026	CVPR 2026; preference alignment personalizes VLA driving behavior to individual styles
2026-03-26	Demographic Fairness in Multimodal LLMs: A Benchmark of Gender and Ethnicity Bias in Face Verification	CVPR 2026	CVPR 2026; demographic fairness benchmark for MLLM-based face verification
2026-03-26	MoE-GRPO: Optimizing Mixture-of-Experts via Reinforcement Learning in Vision-Language Models	CVPR 2026	CVPR 2026; GRPO-based reinforcement learning improves MoE routing in VLMs
2026-03-25	From Weights to Concepts: Data-Free Interpretability of CLIP via Singular Vector Decomposition	CVPR 2026	CVPR 2026; data-free CLIP interpretability via SVD; no activation data required
2026-03-25	Mitigating Object Hallucinations in LVLMs via Attention Imbalance Rectification	CVPR 2026	CVPR 2026; attention imbalance rectification directly mitigates object hallucination
2026-03-25	LensWalk: Agentic Video Understanding by Planning How You See in Videos	CVPR 2026	CVPR 2026; agentic lens planning for dense temporal video understanding
2026-03-25	Unleashing Vision-Language Semantics for Deepfake Video Detection	CVPR 2026	CVPR 2026; VLM semantic priors boost generalizable deepfake video detection
2026-03-24	VISion On Request: Enhanced VLLM efficiency with sparse, dynamically selected, vision-language interactions	CVPR 2026	CVPR 2026; sparse dynamic VL interactions for LVLM efficiency without information bottleneck
2026-03-24	Focus, Don't Prune: Identifying Instruction-Relevant Regions for Information-Rich Image Understanding	CVPR 2026	CVPR 2026; identifies instruction-relevant image regions for information-dense VLM inputs
2026-03-23	Language Models Can Explain Visual Features via Steering	CVPR 2026	CVPR 2026; LLMs explain SAE visual features autonomously via causal steering
2026-03-23	Principled Steering via Null-space Projection for Jailbreak Defense in Vision-Language Models	CVPR 2026	CVPR 2026; null-space projection steering principled defense against visual jailbreaks
2026-03-23	Uncertainty-guided Compositional Alignment with Part-to-Whole Semantic Representativeness in Hyperbolic Vision-Language Models	CVPR 2026	CVPR 2026; hyperbolic embeddings capture part-to-whole hierarchy in VL alignment
2026-03-23	Which Concepts to Forget and How to Refuse? Decomposing Concepts for Continual Unlearning in Large Vision-Language Models	CVPR 2026	CVPR 2026; concept decomposition enables continual selective unlearning in LVLMs
2026-03-22	Uncertainty-Aware Knowledge Distillation for Multimodal Large Language Models	CVPR 2026	CVPR 2026; uncertainty-aware distillation balances teacher vs. data guidance in MLLMs
2026-03-21	Predictive Regularization Against Visual Representation Degradation in Multimodal Large Language Models	CVPR 2026	CVPR 2026; predictive regularization prevents visual representation degradation during LLM training
2026-03-20	When Negation Is a Geometry Problem in Vision-Language Models	CVPR	CVPR; negation failure in CLIP reframed as geometry problem with novel fix
2026-03-20	PersonaVLM: Long-Term Personalized Multimodal LLMs	CVPR 2026	CVPR 2026; multi-turn long-term personalized preference alignment in multimodal LLMs

January 2026 (205)

Date	Paper	Venue	Why selected
2026-01-01	Visual symbolic mechanisms: Emergent symbol processing in Vision Language Models	ICLR 2026	Yoshua Bengio (Mila); ICLR 2026; symbolic binding mechanisms in VLMs
2026-01-01	ThinkMorph: Emergent Properties in Multimodal Interleaved Chain-of-Thought Reasoning	ICLR 2026	Linjie Li (Microsoft); ICLR 2026; multimodal interleaved chain-of-thought emergent properties
2026-01-01	Benchmarking Large Vision-Language Models on Fine-Grained Image Tasks: A Comprehensive Evaluation	ICLR 2026	Serge Belongie (Cornell Tech); ICLR 2026; comprehensive fine-grained LVLM evaluation
2026-01-01	Transductive Visual Programming: Evolving Tool Libraries from Experience for Spatial Reasoning	ICLR 2026	Serena Yeung-Levy (Stanford); ICLR 2026; evolving tool libraries for 3D spatial reasoning
2026-01-01	Scaling up Memory for Robotic Control via Experience Retrieval	ICLR 2026	Chelsea Finn (Stanford); ICLR 2026; scaling memory for robot policies via retrieval
2026-01-01	WorldGym: World Model as An Environment for Policy Evaluation	ICLR 2026	Percy Liang (Stanford); ICLR 2026; world-model as policy evaluation environment
2026-01-01	Cosmos Policy: Fine-Tuning Video Models for Visuomotor Control and Planning	ICLR 2026	Tsung-Yi Lin + Yen-Chen Lin; ICLR 2026; fine-tuning video models for visuomotor control
2026-01-01	HybridVLA: Collaborative Diffusion and Autoregression in a Unified Vision-Language-Action Model	ICLR 2026	Renrui Zhang (CUHK); ICLR 2026; hybrid diffusion+autoregressive VLA model
2026-01-01	InternSpatial: A Comprehensive Dataset for Spatial Reasoning in Vision-Language Models	ICLR 2026	InternVL team (Zhe Chen); ICLR 2026; large-scale spatial reasoning dataset for VLMs
2026-01-01	Discrete Diffusion for Reflective Vision-Language-Action Models in Autonomous Driving	ICLR 2026	Hang Zhao (Tsinghua); ICLR 2026; discrete diffusion VLA for end-to-end autonomous driving
2026-01-01	Map the Flow: Revealing Hidden Pathways of Information in VideoLLMs	ICLR 2026	Bohyung Han (SNU); ICLR 2026; mechanistic analysis of information flow in VideoLLMs
2026-01-01	Zebra-CoT: A Dataset for Interleaved Vision-Language Reasoning	ICLR 2026	ICLR 2026; Zebra-CoT large dataset for interleaved vision-language reasoning
2026-01-01	Self-Improving Vision-Language-Action Models with Data Generation via Residual RL	ICLR 2026	ICLR 2026; self-improving VLA via residual RL, reduces reliance on human demos
2026-01-01	Test-Time Matching: Unlocking Compositional Reasoning in Multimodal Models	ICLR 2026	ICLR 2026; test-time matching unlocks compositional reasoning in frontier VLMs
2026-01-01	Let's Think in Two Steps: Mitigating Agreement Bias in MLLMs with Self-Grounded Verification	ICLR 2026	ICLR 2026; agreement bias mitigation via self-grounded verification in MLLMs
2026-01-01	Importance Sampling for Multi-Negative Multimodal Direct Preference Optimization	ICLR 2026	ICLR 2026; importance-sampling multi-negative multimodal DPO for VLMs
2026-01-01	Hallucination-aware Intermediate Representation Edit in Large Vision-Language Models	ICLR 2026	ICLR 2026; intermediate representation editing for hallucination mitigation
2026-01-01	Dynamic Multimodal Activation Steering for Hallucination Mitigation in Large Vision-Language Models	ICLR 2026	ICLR 2026; dynamic activation steering identifies truthfulness/visual directions
2026-01-01	Matched Data, Better Models: Target Aligned Data Filtering with Sparse Autoencoders	ICLR 2026	ICLR 2026; SAE-based target-aligned data filtering for VLM pretraining
2026-01-01	Label-Free Mitigation of Spurious Correlations in VLMs using Sparse Autoencoders	ICLR 2026	ICLR 2026; label-free SAE-based spurious correlation removal in VLMs
2026-01-01	MetaSpatial: Reinforcing 3D Spatial Reasoning in VLMs for the Metaverse	ICLR 2026	ICLR 2026; first RL framework for 3D spatial reasoning in VLMs for metaverse
2026-01-01	Pursuing Minimal Sufficiency in Spatial Reasoning	ICLR 2026	ICLR 2026; minimal sufficiency framework for spatial reasoning bottlenecks
2026-01-01	Do 3D Large Language Models Really Understand 3D Spatial Relationships?	ICLR 2026	ICLR 2026; exposes 3D LLMs do not truly understand spatial relationships
2026-01-01	SpinBench: Perspective and Rotation as a Lens on Spatial Reasoning in VLMs	ICLR 2026	ICLR 2026; SpinBench: perspective-taking diagnostic for spatial reasoning in VLMs
2026-01-01	ReWatch-R1: Boosting Complex Video Reasoning in Large Vision-Language Models through Agentic Data Synthesis	ICLR 2026	ICLR 2026; RLVR agentic data synthesis for complex video reasoning
2026-01-01	VideoJudge: Bootstrapping Enables Scalable Supervision of MLLM-as-a-Judge for Video Understanding	ICLR 2026	Bhiksha Raj (CMU); ICLR 2026; bootstrapping scalable MLLM-as-judge for video
2026-01-01	VideoReasonBench: Can MLLMs Perform Vision-Centric Complex Video Reasoning?	ICLR 2026	ICLR 2026; VideoReasonBench: vision-centric complex video reasoning benchmark
2026-01-01	FOCUS: Efficient Keyframe Selection for Long Video Understanding	ICLR 2026	ICLR 2026; FOCUS efficient keyframe selection for hour-long video MLLMs
2026-01-01	Memento: Toward an All-Day Proactive Assistant for Ultra-Long Streaming Video	ICLR 2026	ICLR 2026; Memento: proactive assistant for ultra-long streaming video
2026-01-01	SimpleVLA-RL: Scaling VLA Training via Reinforcement Learning	ICLR 2026	ICLR 2026; SimpleVLA-RL: scaling VLA training via reinforcement learning
2026-01-01	X-VLA: Soft-Prompted Transformer as Scalable Cross-Embodiment Vision-Language-Action Model	ICLR 2026	ICLR 2026; X-VLA: soft-prompted transformer for cross-embodiment VLA
2026-01-01	Interleave-VLA: Enhancing Robot Manipulation with Image-Text Interleaved Instructions	ICLR 2026	ICLR 2026; interleaved image-text instructions improve robot manipulation
2026-01-01	PixelVLA: Advancing Pixel-level Understanding in Vision-Language-Action Model	ICLR 2026	ICLR 2026; PixelVLA: pixel-level understanding in vision-language-action models
2026-01-01	HAMLET: Switch Your Vision-Language-Action Model into a History-Aware Policy	ICLR 2026	ICLR 2026; HAMLET: history-aware policy for VLA manipulation models
2026-01-01	TwinVLA: Data-Efficient Bimanual Manipulation with Twin Single-Arm Vision-Language-Action Models	ICLR 2026	ICLR 2026; TwinVLA: data-efficient bimanual manipulation via VLAs
2026-01-01	Embodied-R1: Reinforced Embodied Reasoning for General Robotic Manipulation	ICLR 2026	ICLR 2026; Embodied-R1: unified pointing representation for generalist manipulation
2026-01-01	From Seeing to Doing: Bridging Reasoning and Decision for Robotic Manipulation	ICLR 2026	ICLR 2026; bridging reasoning to action for novel robotic manipulation scenarios
2026-01-01	Hybrid Training for Vision-Language-Action Models	ICLR 2026	ICLR 2026; hybrid VLA training combining CoT with diffusion-based actions
2026-01-01	Unified Diffusion VLA: Vision-Language-Action Model via Joint Discrete Denosing Diffusion Process	ICLR 2026	ICLR 2026; unified discrete diffusion VLA with future image understanding
2026-01-01	Verifier-free Test-Time Sampling for Vision-Language-Action Models	ICLR 2026	ICLR 2026; verifier-free test-time scaling for VLA precision tasks
2026-01-01	Rethinking Bottlenecks in Safety Fine-Tuning of Vision Language Models	ICLR 2026	ICLR 2026; identifies fundamental bottlenecks in VLM safety fine-tuning
2026-01-01	Safety Mirage: How Spurious Correlations Undermine VLM Safety Fine-Tuning and Can Be Mitigated by Machine Unlearning	ICLR 2026	ICLR 2026; spurious correlations undermine safety fine-tuning; machine unlearning fix
2026-01-01	ARMs: Adaptive Red-Teaming Agent against Multimodal Models with Plug-and-Play Attacks	ICLR 2026	ICLR 2026; adaptive plug-and-play red-teaming agents against multimodal models
2026-01-01	Misaligned Roles, Misplaced Images: Structural Input Perturbations Expose Multimodal Alignment Blind Spots	ICLR 2026	ICLR 2026; structural input perturbations reveal multimodal alignment blind spots
2026-01-01	AdPO: Enhancing the Adversarial Robustness of Large Vision-Language Models with Preference Optimization	ICLR 2026	ICLR 2026; AdPO: preference optimization for adversarial robustness in LVLMs
2026-01-01	BEAT: Visual Backdoor Attacks on VLM-based Embodied Agents via Contrastive Trigger Learning	ICLR 2026	ICLR 2026; BEAT: contrastive visual backdoor attacks on VLM embodied agents
2026-01-01	Do Vision-Language Models Respect Contextual Integrity in Location Disclosure?	ICLR 2026	ICLR 2026; VLMs violate contextual integrity in location privacy disclosure
2026-01-01	TrustGen: A Platform of Dynamic Benchmarking on the Trustworthiness of Generative Foundation Models	ICLR 2026	ICLR 2026; TrustGen: dynamic benchmarking platform for generative model trustworthiness
2026-01-01	Kaleidoscope: In-language Exams for Massively Multilingual Vision Evaluation	ICLR 2026	ICLR 2026; Kaleidoscope: massively multilingual in-language VLM evaluation
2026-01-01	RAVENEA: A Benchmark for Multimodal Retrieval-Augmented Visual Culture Understanding	ICLR 2026	ICLR 2026; RAVENEA: retrieval-augmented visual culture understanding benchmark
2026-01-01	Vision Language Models are Biased	ICLR 2026	ICLR 2026; systematic evidence that VLMs exhibit strong prior-knowledge biases
2026-01-01	MMReD: a Cross-Modal Benchmark for Dense Context Reasoning	ICLR 2026	ICLR 2026; MMReD: dense multi-modal context reasoning over long inputs
2026-01-01	ChartGalaxy: A Dataset for Infographic Chart Understanding and Generation	ICLR 2026	ICLR 2026; ChartGalaxy: large-scale infographic chart understanding and generation
2026-01-01	GeoBench: Rethinking Multimodal Geometric Problem-Solving via Hierarchical Evaluation	ICLR 2026	ICLR 2026; GeoBench: hierarchical geometric reasoning evaluation for VLMs
2026-01-01	Unified Vision–Language Modeling via Concept Space Alignment	ICLR 2026	ICLR 2026; V-SONAR: extends multilingual text embeddings to 1500-language vision
2026-01-01	VisCodex: Unified Multimodal Code Generation via Merging Vision and Coding Models	ICLR 2026	ICLR 2026; VisCodex: unified multimodal code generation merging vision and code models
2026-01-01	DiVE-k: DIFFERENTIAL VISUAL REASONING FOR FINE-GRAINED IMAGE RECOGNITION	ICLR 2026	Ram Nevatia (USC); ICLR 2026; differential visual reasoning for fine-grained recognition
2026-01-01	Sparse CLIP: Co-Optimizing Interpretability and Performance in Contrastive Learning	ICLR 2026	ICLR 2026; Sparse CLIP co-optimizes interpretability and performance in CLIP
2026-01-01	Post-hoc Probabilistic Vision-Language Models	ICLR 2026	ICLR 2026; post-hoc probabilistic VLMs with calibrated uncertainty in joint space
2026-01-01	Small Drafts, Big Verdict: Information-Intensive Visual Reasoning via Speculation	ICLR 2026	ICLR 2026; speculative drafting approach for information-intensive visual reasoning
2026-01-01	Game-RL: Synthesizing Multimodal Verifiable Game Data to Boost VLMs' General Reasoning	ICLR 2026	ICLR 2026; game-based multimodal verifiable data boosts VLM general reasoning
2026-01-01	ProxyThinker: Test-Time Guidance through Small Visual Reasoners	ICLR 2026	ICLR 2026; small proxy VLMs guide test-time reasoning of large VLMs cheaply
2026-01-01	Vision-SR1: Self-Rewarding Vision-Language Model via Reasoning Decomposition and Multi-Reward Policy Optimization	ICLR 2026	ICLR 2026; Vision-SR1: self-rewarding VLM via multi-reward policy optimization
2026-01-01	ViPER: Empowering the Self-Evolution of Visual Perception Abilities in Vision-Language Models	ICLR 2026	ICLR 2026; ViPER: self-evolution for fine-grained visual perception in VLMs
2026-01-01	Vision-Zero: Scalable VLM Self-Evolution via Multi-Agent Self-Play	ICLR 2026	ICLR 2026; Vision-Zero: multi-agent self-play for label-free VLM self-evolution
2026-01-01	CompoDistill: Attention Distillation for Compositional Reasoning in Multimodal LLMs	ICLR 2026	ICLR 2026; attention distillation from large to small MLLMs for compositional reasoning
2026-01-01	Imitating the Truth: Attention-aware Truth-Guided Enhancement for Hallucination Mitigation in Large Vision-Language Models	ICLR 2026	ICLR 2026; attention-aware truth-guided enhancement for LVLM hallucination
2026-01-01	Seeing but Not Believing: Probing the Disconnect Between Visual Attention and Answer Correctness in VLMs	ICLR 2026	ICLR 2026; probes disconnect between visual attention and answer correctness in VLMs
2026-01-01	AgilePruner: An Empirical Study of Attention and Diversity for Adaptive Visual Token Pruning in Large Vision-Language Models	ICLR 2026	ICLR 2026; empirical study of attention+diversity for adaptive visual token pruning
2026-01-01	LearnPruner: Rethinking Attention-based Token Pruning in Vision Language Models	ICLR 2026	ICLR 2026; rethinks attention-based token pruning in VLMs
2026-01-01	Prune Redundancy, Preserve Essence: Vision Token Compression in VLMs via Synergistic Importance-Diversity	ICLR 2026	ICLR 2026; synergistic importance-diversity token compression for VLMs
2026-01-01	WSVD: Weighted Low-Rank Approximation for Fast and Efficient Execution of Low-Precision Vision-Language Models	ICLR 2026	ICLR 2026; weighted SVD for fast low-precision VLM execution
2026-01-01	Embodied Navigation Foundation Model	ICLR 2026	ICLR 2026; foundation model for embodied navigation leveraging VLM priors
2026-01-01	CitySeeker: How Do VLMs Explore Embodied Urban Navigation with Implicit Human Needs?	ICLR 2026	ICLR 2026; VLMs interpreting implicit human needs in urban embodied navigation
2026-01-01	Talking Points: Describing and Localizing Pixels	ICLR 2026	ICLR 2026; Talking Points: pixel-precise keypoint comprehension via natural language
2026-01-01	Decomposition of Concept-Level Rules in Visual Scenes	ICLR 2026	ICLR 2026; decomposing concept-level visual rules for compositional scene parsing
2026-01-01	DaVinci: Reinforcing Visual-Structural Syntax in MLLMs for Generalized Scientific Diagram Parsing	ICLR 2026	ICLR 2026; DaVinci: RL-driven visual-structural syntax for scientific diagram parsing
2026-01-01	Mordal: Automated Pretrained Model Selection for Vision Language Models	ICLR 2026	ICLR 2026; Mordal: automated pretrained model selection for building VLMs
2026-01-01	EditReward: A Human-Aligned Reward Model for Instruction-Guided Image Editing	ICLR 2026	ICLR 2026; EditReward: human-aligned reward model for instruction-guided image editing
2026-01-01	VLMgineer: Vision-Language Models as Robotic Toolsmiths	ICLR 2026	ICLR 2026; VLMs as robotic toolsmiths — tool design as measure of physical intelligence
2026-01-01	VITA: Zero-Shot Value Functions via Test-Time Adaptation of Vision–Language Models	ICLR 2026	ICLR 2026; VITA: test-time adaptation of VLMs as zero-shot goal-conditioned value functions
2026-01-01	DeepEyes: Incentivizing "Thinking with Images" via Reinforcement Learning	ICLR 2026	RL incentivizes visual thinking in VLMs; novel image-based reasoning paradigm
2026-01-01	VTool-R1: VLMs Learn to Think with Images via Reinforcement Learning on Multimodal Tool Use	ICLR 2026	VLMs learn multimodal tool use via RL; key agentic VLM capability
2026-01-01	Revisiting Multimodal Positional Encoding in Vision–Language Models	ICLR 2026	Comprehensive multimodal RoPE analysis; foundational for VLM positional encoding
2026-01-01	To Sink or Not to Sink: Visual Information Pathways in Large Vision-Language Models	ICLR 2026	Mechanistic analysis of visual information pathways and attention sinks in LVLMs
2026-01-01	Generative Universal Verifier as Multimodal Meta-Reasoner	ICLR 2026	Generative universal verifier enables VLM meta-reasoning and self-refinement
2026-01-01	Cross-Modal Redundancy and the Geometry of Vision–Language Embeddings	ICLR 2026	Geometric probe of VL joint embedding space via cross-modal redundancy
2026-01-01	OneTwoVLA: A Unified Vision-Language-Action Model with Adaptive Reasoning	ICLR 2026	OneTwoVLA: unified VLA with adaptive reasoning-acting switching for robotics
2026-01-01	Actions as Language: Fine-Tuning VLMs into VLAs Without Catastrophic Forgetting	ICLR 2026	Fine-tuning VLMs into VLAs without catastrophic forgetting; solves core challenge
2026-01-01	Rethinking Causal Mask Attention for Vision-Language Inference	ICLR 2026	Rethinks causal masking in autoregressive VLMs; non-trivial architectural insight
2026-01-01	PPE: Positional Preservation Embedding for Token Compression in Multimodal Large Language Models	ICLR 2026	Positional information preserved during token merging prevents spatial degradation
2026-01-01	GuardAlign: Test-time Safety Alignment in Multimodal Large Language Models	ICLR 2026	Test-time safety alignment for LVLMs; input detection plus output steering
2026-01-01	Toward Universal and Transferable Jailbreak Attacks on Vision-Language Models	ICLR 2026	Universal transferable jailbreak attacks on VLMs; multimodal attack surface analysis
2026-01-01	From Sure" toSorry": Detecting Jailbreak in Large Vision Language Model via JailNeurons	ICLR 2026	JailNeurons: fast neuron-based jailbreak detection for LVLMs; ICLR 2026
2026-01-01	Dyslexify: A Mechanistic Defense Against Typographic Attacks in CLIP	ICLR 2026	Mechanistic defense against typographic attacks in CLIP; ICLR 2026 analysis
2026-01-01	Identifying Robust Neural Pathways: Few-Shot Adversarial Mask Tuning for Vision-Language Models	ICLR 2026	Adversarial mask tuning identifies robust neural pathways in VLMs
2026-01-01	Get RICH or Die Scaling: Profitably Trading Inference Compute for Robustness	ICLR 2026	Inference-time compute improves OOD robustness in VLMs; ICLR 2026 analysis
2026-01-01	Spatial Reasoning with Vision-Language Models in Ego-Centric Multi-View Scenes	ICLR 2026	Egocentric multi-view spatial reasoning benchmark; VLMs struggle with 3D
2026-01-01	SpatiaLab: Can Vision–Language Models Perform Spatial Reasoning in the Wild?	ICLR 2026	In-the-wild spatial reasoning benchmark challenges current VLM spatial capabilities
2026-01-01	OmniSpatial: Towards Comprehensive Spatial Reasoning Benchmark for Vision Language Models	ICLR 2026	Comprehensive 3D spatial reasoning benchmark with diverse relationship categories
2026-01-01	IndicVisionBench: Benchmarking Cultural and Multilingual Understanding in VLMs	ICLR 2026	South-Asian cultural VLM benchmark exposes Western-centric evaluation blindspots
2026-01-01	Contamination Detection for VLMs Using Multi‑Modal Semantic Perturbations	ICLR 2026	Multi-modal semantic perturbations detect VLM benchmark data contamination
2026-01-01	IV-Bench: A Benchmark for Image-Grounded Video Perception and Reasoning in Multimodal LLMs	ICLR 2026	Image-grounded video benchmark tests VLMs on cross-modal video comprehension
2026-01-01	Spotlight on Token Perception for Multimodal Reinforcement Learning	ICLR 2026	Token-level perceptual spotlight improves RLVR visual reasoning in VLMs
2026-01-01	SPECS: Decoupling Multimodal Learning via Self-distilled Preference-based Cold Start	ICLR 2026	Self-distilled cold start decouples visual/language learning for RLVR VLMs
2026-01-01	P\(^2\)-DPO: Grounding Hallucination in Perceptual Processing via Calibration Direct Preference Optimization	ICLR 2026	Calibration DPO grounds hallucination in perceptual processing; preference learning
2026-01-01	Action-aware Dynamic Pruning for Efficient Vision-Language-Action Manipulation	ICLR 2026	Action-aware dynamic pruning enables efficient VLA inference; ICLR 2026
2026-01-01	villa-X: Enhancing Latent Action Modeling in Vision-Language-Action Models	ICLR 2026	villa-X: enhanced latent action modeling for VLA models; architectural advance
2026-01-01	Robust Fine-tuning of Vision-Language-Action Robot Policies via Parameter Merging	ICLR 2026	Parameter merging preserves generalist VLA skills during task-specific fine-tuning
2026-01-01	Memory-Free Continual Learning with Null Space Adaptation for Zero-Shot Vision-Language Models	ICLR 2026	Null space adaptation enables memory-free continual VLM learning; preserves zero-shot
2026-01-01	Enhanced Continual Learning of Vision-Language Models with Model Fusion	ICLR 2026	Model fusion enhances continual VLM learning; mitigates catastrophic forgetting
2026-01-01	KeepLoRA: Continual Learning with Residual Gradient Adaptation	ICLR 2026	KeepLoRA: residual gradient adaptation balances VLM plasticity and stability
2026-01-01	Bongard-RWR+: Real-World Representations of Fine-Grained Concepts in Bongard Problems	ICLR 2026	Real-world Bongard problems test abstract visual concept learning in VLMs
2026-01-01	PoSh: Using Scene Graphs to Guide LLMs-as-a-Judge for Detailed Image Descriptions	ICLR 2026	Scene graph-guided LLM-as-judge improves detailed image description evaluation
2026-01-01	K-Sort Eval: Efficient Preference Evaluation for Visual Generation via Corrected VLM-as-a-Judge	ICLR 2026	K-Sort corrects VLM-as-judge bias; scalable visual generation preference evaluation
2026-01-01	Evaluating Cross-Modal Reasoning Ability and Problem Characteristics with Multimodal Item Response Theory	ICLR 2026	Multimodal Item Response Theory enables rigorous VLM reasoning benchmark design
2026-01-01	DeCo-DETR: Decoupled Cognition DETR for efficient Open-Vocabulary Object Detection	ICLR 2026	DeCo-DETR decoupled cognition achieves efficient open-vocabulary detection; ICLR 2026
2026-01-01	Fantastic Tractor-Dogs and How Not to Find Them With Open-Vocabulary Detectors	ICLR 2026	OVDs produce confident false positives on negative images; critical deployment flaw
2026-01-01	VideoZoomer: Reinforcement-Learned Temporal Focusing for Long Video Reasoning	ICLR 2026	RL temporal focusing on relevant frames enables efficient long-video VLM reasoning
2026-01-01	CaReBench: A Fine-grained Benchmark for Video Captioning and Retrieval	ICLR 2026	Fine-grained video captioning and retrieval benchmark reveals VLM temporal limits
2026-01-01	STVG-R1: Incentivizing Instance-Level Reasoning and Grounding in Videos via Reinforcement Learning	ICLR 2026	RL reduces hallucination in spatial-temporal video grounding; ICLR 2026
2026-01-01	Enhancing Multi-Image Understanding through Delimiter Token Scaling	ICLR 2026	Delimiter token scaling resolves cross-image leakage in multi-image VLMs
2026-01-01	QLIP: A Dynamic Quadtree Vision Prior Enhances MLLM Performance Without Retraining	ICLR 2026	Quadtree spatial prior boosts MLLM performance without retraining; ICLR 2026
2026-01-01	ERGO: Efficient High-Resolution Visual Understanding for Vision-Language Models	ICLR 2026	ERGO efficient high-resolution VLM processing via selective token allocation
2026-01-01	MobileIPL: Enhancing Mobile Agents Thinking Process via Iterative Preference Learning	ICLR 2026	Iterative preference learning improves mobile VLM agent chain-of-action reasoning
2026-01-01	SMAN-Bench: A Cross-System Benchmark for Mobile Agents under Single- and Multi-path, Ambiguous, and Noisy Tasks	ICLR 2026	Cross-system mobile agent benchmark with ambiguous and noisy real-world tasks
2026-01-01	Simulation to Rules: A Dual-VLM Framework for Formal Visual Planning	ICLR 2026	Dual-VLM bridges formal PDDL planning with visual scene understanding
2026-01-01	DriveAgent-R1: Advancing VLM-based Autonomous Driving with Active Perception and Hybrid Thinking	ICLR 2026	DriveAgent-R1: active perception and hybrid thinking for autonomous driving VLMs
2026-01-01	Detecting Misbehaviors of Large Vision-Language Models by Evidential Uncertainty Quantification	ICLR 2026	Evidential uncertainty quantification detects VLM misbehaviors on shifted inputs
2026-01-01	Revisiting Confidence Calibration for Misclassification Detection in VLMs	ICLR 2026	Standard calibration degrades misclassification detection in VLMs; key theoretical finding
2026-01-01	Asynchronous Matching with Dynamic Sampling for Multimodal Dataset Distillation	ICLR 2026	Asynchronous matching for multimodal dataset distillation; efficient VLM training
2026-01-01	Ref-Adv: Exploring MLLM Visual Reasoning in Referring Expression Tasks	ICLR 2026	Adversarial referring expression tasks expose gaps in VLM visual reasoning
2026-01-01	WIMFRIS: WIndow Mamba Fusion and Parameter Efficient Tuning for Referring Image Segmentation	ICLR 2026	Window Mamba fusion with parameter-efficient tuning for referring segmentation
2026-01-01	LINK: Learning Instance-level Knowledge from Vision-Language Models for Human-Object Interaction Detection	ICLR 2026	Instance-level VLM knowledge for HOI detection; balances generalization and specialization
2026-01-01	WAVE: Learning Unified & Versatile Audio-Visual Embeddings with Multimodal LLM	ICLR 2026	WAVE: unified audio-visual VLM embeddings outperform dedicated audio-visual pretraining
2026-01-01	Modality Alignment across Trees on Heterogeneous Hyperbolic Manifolds	ICLR 2026	Hyperbolic manifold alignment for hierarchical multimodal feature fusion in VLMs
2026-01-01	MoRA: Missing Modality Low-Rank Adaptation for Visual Recognition	ICLR 2026	Low-rank adaptation handles missing modalities in VLM visual recognition tasks
2026-01-01	Guided Query Refinement: Multimodal Hybrid Retrieval with Test-Time Optimization	ICLR 2026	Test-time query refinement bridges modality gap in multimodal retrieval with VLMs
2026-01-01	DAVE: A VLM Vision Encoder for Document Understanding and Web Agents	ICLR 2026	DAVE encoder captures low-level spatial structure improving document VLM performance
2026-01-01	Adaptive Logit Adjustment for Debiasing Multimodal Language Models	ICLR 2026	Adaptive logit adjustment debiases VLM captioning and VQA generation tasks
2026-01-01	WRING Out The Bias: A Rotation-Based Alternative To Projection Debiasing	ICLR 2026	Rotation-based debiasing removes CLIP spurious correlations; outperforms projection
2026-01-01	Read the Room: Video Social Reasoning with Mental-Physical Causal Chains	ICLR 2026	Video social reasoning benchmark with mental-physical causal chains; VLM ToM
2026-01-01	ChainMPQ: Interleaved Text-Image Reasoning Chains for Mitigating Relation Hallucinations	ICLR 2026	Interleaved text-image reasoning chains mitigate relation hallucination in VLMs
2026-01-01	Hallucination Reduction with CASAL: Contrastive Activation Steering for Amortized Learning	ICLR 2026	Contrastive activation steering reduces hallucination via linear knowledge representations
2026-01-01	Reasoning-Driven Multimodal LLM for Domain Generalization	ICLR 2026	MLLM reasoning improves domain generalization beyond visual feature invariance
2026-01-01	Discrete Latent Features Ablate Adversarial Attack: A Robust Prompt Tuning Framework for VLMs	ICLR 2026	Discrete latent features resist adversarial attacks in VLM prompt tuning
2026-01-01	Inducing Dyslexia in Vision Language Models	ICLR 2026	Mechanistic VLM study reveals VWFA analog in CLIP vision encoders
2026-01-01	AQuA: Toward Strategic Response Generation for Ambiguous Visual Questions	ICLR 2026	Strategic response generation for ambiguous VQA; addresses real-world uncertainty
2026-01-01	VL-JEPA: Joint Embedding Predictive Architecture for Vision-language	ICLR 2026	Novel JEPA architecture for VLMs; continuous embedding prediction vs. autoregressive tokens
2026-01-01	No Labels, No Problem: Training Visual Reasoners with Multimodal Verifiers	ICLR 2026	Georgia Gkioxari/Meta; label-free visual reasoning training via multimodal verifiers
2026-01-01	Unleashing Perception-Time Scaling to Multimodal Reasoning Models	ICLR 2026	Extends inference-time compute scaling to VLM perception; frontier capability paper
2026-01-01	PostAlign: Multimodal Grounding as a Corrective Lens for MLLMs	ICLR 2026	Philip Torr/Oxford; multimodal grounding corrects MLLM fine-grained visual failures
2026-01-01	MaskInversion: Localized Embeddings via Optimization of Explainability Maps	ICLR 2026	Ferrari/Rupprecht (Google/Oxford); data-free localized CLIP embeddings via mask optimization
2026-01-01	RegionReasoner: Region-Grounded Multi-Round Visual Reasoning	ICLR 2026	Cees Snoek; iterative region-grounded multi-round visual reasoning beyond single-pass VLMs
2026-01-01	Understanding Language Prior of LVLMs by Contrasting Chain-of-Embedding	ICLR 2026	ICLR 2026; chain-of-embedding contrast exposes language prior dominance in LVLMs
2026-01-01	LLaVA-FA: Learning Fourier Approximation for Compressing Large Multimodal Models	ICLR 2026	ICLR 2026; Fourier approximation compresses large multimodal models with strong efficiency
2026-01-01	Mitigating Hallucination in Vision-Language Model with Depth and Spatial-aware Key-Value Refinement	ICLR 2026	ICLR 2026; depth and spatial-aware KV cache refinement reduces VLM hallucination
2026-01-01	AFTER: Mitigating the Object Hallucination of LVLM via Adaptive Factual-Guided Activation Editing	ICLR 2026	ICLR 2026; adaptive factual-guided activation editing targets object hallucination categories
2026-01-01	Self-Aug: Query and Entropy Adaptive Decoding for Large Vision-Language Models	ICLR 2026	ICLR 2026; query-adaptive entropy-based contrastive decoding mitigates LVLM hallucination
2026-01-01	Cat-PO: Cross-modal Adaptive Token-rewards for Preference Optimization in Truthful Multimodal LLMs	ICLR 2026	ICLR 2026; cross-modal token-level reward optimization reduces MLLM hallucination
2026-01-01	GHOST: Hallucination-Inducing Image Generation for Multimodal LLMs	ICLR 2026	ICLR 2026; adversarial image synthesis stress-tests MLLM hallucination failure modes
2026-01-01	Transferable and Stealthy Adversarial Attacks on Large Vision-Language Models	ICLR 2026	ICLR 2026; progressive semantic infusion yields transferable stealthy LVLM attacks
2026-01-01	Seeing What’s Not There: Negation Understanding Needs More Than Training	ICLR 2026	ICLR 2026; shows negation understanding requires more than training data scaling
2026-01-01	NePTune: A Neuro-Pythonic Framework for Tunable Compositional Reasoning on Vision-Language	ICLR 2026	ICLR 2026; neuro-symbolic tunable compositional reasoning framework for VLMs
2026-01-01	Distributional Vision-Language Alignment by Cauchy-Schwarz Divergence	ICLR 2026	ICLR 2026; Cauchy-Schwarz divergence for distributional VL alignment beyond pairwise InfoNCE
2026-01-01	Spatial-DISE: A Unified Benchmark for Evaluating Spatial Reasoning in Vision-Language Models	ICLR 2026	ICLR 2026; unified spatial reasoning benchmark spanning robotics, AR, and navigation
2026-01-01	TABLET: A Large-Scale Dataset for Robust Visual Table Understanding	ICLR 2026	Mirella Lapata; large-scale real-world dataset for robust visual table understanding
2026-01-01	FrameThinker: Learning to Think with Long Videos via Multi-Turn Frame Spotlighting	ICLR 2026	ICLR 2026; multi-turn frame spotlighting for efficient long-video VLM reasoning
2026-01-01	V2P-Bench: Evaluating Video-Language Understanding with Visual Prompts for Better Human-Model Interaction	ICLR 2026	ICLR 2026; visual prompt-based video benchmark enables richer human-model interaction
2026-01-01	A Training-Free Framework for Long Video Understanding via Video-Query-Options Similarity	ICLR 2026	ICLR 2026; training-free long-video understanding via video-query-options similarity
2026-01-01	WMPO: World Model-based Policy Optimization for Vision-Language-Action Models	ICLR 2026	ICLR 2026; world-model RL policy optimization for VLA; goes beyond demonstrations
2026-01-01	On Robustness of Vision-Language-Action Model against Multi-Modal Perturbations	ICLR 2026	ICLR 2026; systematic study of VLA robustness to multi-modal perturbations
2026-01-01	Sim2Real VLA: Zero-Shot Generalization of Synthesized Skills to Realistic Manipulation	ICLR 2026	ICLR 2026; zero-shot sim-to-real VLA transfer via synthesized skill generalization
2026-01-01	Align-Then-stEer: Adapting the Vision-Language Action Models through Unified Latent Guidance	ICLR 2026	ICLR 2026; unified latent guidance adapts VLA models to downstream manipulation tasks
2026-01-01	QVLA: Not All Channels Are Equal in Vision-Language-Action Model's Quantization	ICLR 2026	ICLR 2026; channel-aware quantization enables VLA deployment on resource-constrained robots
2026-01-01	EVLP: Learning Unified Embodied Vision-Language Planner with Reinforced Supervised Fine-Tuning	ICLR 2026	ICLR 2026; reinforced supervised fine-tuning for unified embodied VL planner
2026-01-01	WebWatcher: Breaking New Frontiers of Vision-Language Deep Research Agent	ICLR 2026	ICLR 2026; VL deep research agent navigates visual web for complex information-seeking
2026-01-01	GhostEI-Bench: Do Mobile Agent Resilience to Environmental Injection in Dynamic On-Device Environments?	ICLR 2026	ICLR 2026; benchmark for VLM agent resilience to environmental injection on mobile
2026-01-01	Uncertainty-Aware Gaussian Map for Vision-Language Navigation	ICLR 2026	ICLR 2026; uncertainty-aware Gaussian map improves VLN in partially observed 3D environments
2026-01-01	Ground Slow, Move Fast: A Dual-System Foundation Model for Generalizable Vision-Language Navigation	ICLR 2026	ICLR 2026; dual-system slow-grounding/fast-moving VLM for generalizable VLN
2026-01-01	Beyond Classification Accuracy: Neural-MedBench and the Need for Deeper Reasoning Benchmarks	ICLR 2026	ICLR 2026; argues current medical VLM benchmarks mask true clinical reasoning gaps
2026-01-01	TumorChain: Interleaved Multimodal Chain-of-Thought Reasoning for Traceable Clinical Tumor Analysis	ICLR 2026	ICLR 2026; interleaved multimodal CoT enables traceable clinical tumor analysis
2026-01-01	MedVR: Annotation-Free Medical Visual Reasoning via Agentic Reinforcement Learning	ICLR 2026	ICLR 2026; annotation-free medical VLM reasoning via agentic reinforcement learning
2026-01-01	OmniCT: Towards a Unified Slice-Volume LVLM for Comprehensive CT Analysis	ICLR 2026	ICLR 2026; unified slice-volume LVLM covering full CT image analysis pipeline
2026-01-01	MMedAgent-RL: Optimizing Multi-Agent Collaboration for Multimodal Medical Reasoning	ICLR 2026	ICLR 2026; multi-agent RL framework for specialist medical VLM collaboration
2026-01-01	Towards Text-Mask Consistency in Medical Image Segmentation	ICLR 2026	ICLR 2026; text-mask consistency constraints fix VLM multi-lesion segmentation failures
2026-01-01	Reversible Primitive–Composition Alignment for Continual Vision–Language Learning	ICLR 2026	ICLR 2026; reversible primitive-composition alignment for tight-budget continual VL learning
2026-01-01	Pi-CCA: Prompt-Invariant CCA Certificates for Replay-Free Continual Multimodal Learning	ICLR 2026	ICLR 2026; prompt-invariant CCA certificates enable replay-free continual VL learning
2026-01-01	RLAP-CLIP: Continual Multimodal Learning with Prototype Adaptation and Difficulty-Aware Routing	ICLR 2026	ICLR 2026; prototype adaptation with difficulty routing for class-incremental CLIP
2026-01-01	Naming to Learn: Class Incremental Learning for Vision-Language Model with Unlabeled Data	ICLR 2026	ICLR 2026; naming paradigm enables class incremental VLM learning from unlabeled data
2026-01-01	A-TPT: Angular Diversity Calibration Properties for Test-Time Prompt Tuning of Vision-Language Models	ICLR 2026	ICLR 2026; angular diversity calibration addresses prompt dispersion gap in VLM TTA
2026-01-01	Flatness Guided Test-Time Adaptation for Vision-Language Models	ICLR 2026	ICLR 2026; flatness-guided TTA links loss landscape geometry to VLM distribution shift
2026-01-01	Bilateral Information-aware Test-time Adaptation for Vision-Language Models	ICLR 2026	ICLR 2026; bilateral information-aware TTA handles covariate shifts without test distribution
2026-01-01	Adaptive Debiasing Tsallis Entropy for Test-Time Adaptation	ICLR 2026	ICLR 2026; Tsallis entropy debiasing corrects CLIP's built-in bias during TTA
2026-01-01	Benchmarking Bias Mitigation Toward Fairness Without Harm from Vision to LVLMs	ICLR 2026	ICLR 2026; first comprehensive benchmark comparing bias mitigation in VLMs
2026-01-01	Semantic Visual Anomaly Detection and Reasoning in AI-Generated Images	ICLR 2026	ICLR 2026; VLM-based semantic anomaly detection in AI-generated image outputs
2026-01-01	Towards Faithful Reasoning in Remote Sensing: A Perceptually-Grounded GeoSpatial Chain-of-Thought for Vision-Language Models	ICLR 2026	ICLR 2026; perceptually grounded chain-of-thought for faithful VLM geospatial reasoning
2026-01-01	FLUX-Reason-6M & PRISM-Bench: A Million-Scale Text-to-Image Reasoning Dataset and Comprehensive Benchmark	ICLR 2026	ICLR 2026; million-scale T2I reasoning dataset and benchmark from top labs
2026-01-01	Object-Centric Refinement for Enhanced Zero-Shot Segmentation	ICLR 2026	ICLR 2026; object-centric region refinement enhances CLIP zero-shot segmentation
2026-01-01	Zero-shot HOI Detection with MLLM-based Detector-agnostic Interaction Recognition	ICLR 2026	ICLR 2026; MLLM as detector-agnostic interaction recognizer for zero-shot HOI
2026-01-01	lmgame-Bench: How Good are LLMs at Playing Games?	ICLR 2026	Eric P. Xing/CMU; game-playing benchmark requiring perception, reasoning, planning
2026-01-01	pFedMMA: Personalized Federated Fine-Tuning with Multi-Modal Adapter for Vision-Language Models	ICLR 2026	ICLR 2026; personalized federated fine-tuning for CLIP with multi-modal adapters
2026-01-01	Self-Evolving Vision-Language Models for Image Quality Assessment via Voting and Ranking	ICLR 2026	ICLR 2026; self-supervised voting and ranking evolves VLMs for image quality assessment
2026-01-01	VLBiMan: Vision-Language Anchored One-Shot Demonstration Enables Generalizable Bimanual Robotic Manipulation	ICLR 2026	ICLR 2026; VL anchoring enables one-shot generalization for bimanual manipulation
2026-01-01	Seeing What’s Wrong: A Trajectory-Guided Approach to Caption Error Detection	ICLR 2026	ICLR 2026; trajectory-guided multi-score method detects subtle image-caption errors