ICLR 2026 accepted papers on Vision-language and multimodal
555 of the 5,358 papers accepted at ICLR 2026 (10.4%) are on Vision-language and multimodal, compared with 7.3% at ICLR 2025. They fall into 6 topics.
See ICLR 2026 in the interactive map
Working on something in this area? Describe your idea in scime atlas to see the closest accepted papers by meaning and how crowded the spot has become.
Topics and papers
Multimodal LLMs · 323 papers
- MedAgent-Pro: Towards Evidence-based Multi-modal Medical Diagnosis via Reasoning Agentic Workflow (13 citations)
- Think Then Embed: Generative Context Improves Multimodal Embedding (2 citations)
- cadrille: Multi-modal CAD Reconstruction with Reinforcement Learning (1 citation)
- FRIEDA: Benchmarking Multi-Step Cartographic Reasoning in Vision-Language Models (1 citation)
- Seeing Across Views: Benchmarking Spatial Reasoning of Vision-Language Models in Robotic Scenes (1 citation)
- SpatialViz-Bench: A Cognitively-Grounded Benchmark for Diagnosing Spatial Visualization in MLLMs (1 citation)
- Understanding Language Prior of LVLMs by Contrasting Chain-of-Embedding (1 citation)
- Vision Language Models are Biased (1 citation)
- Vision-R1: Incentivizing Reasoning Capability in Multimodal Large Language Models (1 citation)
- XModBench: Benchmarking Cross-Modal Capabilities and Consistency in Omni-Language Models (1 citation)
- $AutoDrive\text{-}P^3$: Unified Chain of Perception–Prediction–Planning Thought via Reinforcement Fine-Tuning
- A Comprehensive Information-Decomposition Analysis of Large Vision-Language Models
and 311 more in the interactive map.
Robot action and navigation · 76 papers
- OneTwoVLA: A Unified Vision-Language-Action Model with Adaptive Reasoning (1 citation)
- ReCogDrive: A Reinforced Cognitive Framework for End-to-End Autonomous Driving (1 citation)
- Align-Then-stEer: Adapting the Vision-Language Action Models through Unified Latent Guidance
- All-day Multi-scenes Lifelong Vision-and-Language Navigation with Tucker Adaptation
- AnyTouch 2: General Optical Tactile Representation Learning For Dynamic Tactile Perception
- AutoBio: A Simulation and Benchmark for Robotic Automation in Digital Biology Laboratory
- AutoFly: Vision-Language-Action Model for UAV Autonomous Navigation in the Wild
- Cortical Policy: A Dual-Stream View Transformer for Robotic Manipulation
- D2E: Scaling Vision-Action Pretraining on Desktop Data for Transfer to Embodied AI
- Disentangled Robot Learning via Separate Forward and Inverse Dynamics Pretraining
- DriveVLA-W0: World Models Amplify Data Scaling Law in Autonomous Driving
- Embodied-R1: Reinforced Embodied Reasoning for General Robotic Manipulation
and 64 more in the interactive map.
Cross-modal learning · 59 papers
- WAVE: Learning Unified & Versatile Audio-Visual Embeddings with Multimodal LLM (2 citations)
- BioX-Bridge: Model Bridging for Unsupervised Cross-Modal Knowledge Transfer across Biosignals
- Cross-Modal Redundancy and the Geometry of Vision–Language Embeddings
- Delving into Spectral Clustering with Vision-Language Representations
- Distributional Vision-Language Alignment by Cauchy-Schwarz Divergence
- Efficient Audio-Visual Speech Separation with Discrete Lip Semantics and Multi-Scale Global-Local Attention
- FLARE: Fully Integration of Vision-Language Representations for Deep Cross-Modal Understanding
- FlowBind: Efficient Any-to-Any Generation with Bidirectional Flows
- From Natural Alignment to Conditional Controllability in Multimodal Dialogue
- Fusing Pixels and Genes: Spatially-Aware Learning in Computational Pathology
- Guided Query Refinement: Multimodal Hybrid Retrieval with Test-Time Optimization
- ICYM2I: The illusion of multimodal informativeness under missingness
and 47 more in the interactive map.
Human and embodied modeling · 44 papers
- Into the Rabbit Hull: From Task-Relevant Concepts in DINO to Minkowski Geometry (5 citations)
- VL-JEPA: Joint Embedding Predictive Architecture for Vision-language (1 citation)
- A Cognitive Process-Inspired Architecture for Subject-Agnostic Brain Visual Decoding
- Assembling the Mind's Mosaic: Towards EEG Semantic Intent Decoding
- AttTok: Marrying Attribute Tokens with Generative Pre-trained Vision-Language Models towards Medical Image Understanding
- BioCAP: Exploiting Synthetic Captions Beyond Labels in Biological Foundation Models
- Generate Any Scene: Scene Graph Driven Data Synthesis for Visual Generation Training
- HierLoc: Hyperbolic Entity Embeddings for Hierarchical Visual Geolocation
- Image Quality Assessment for Embodied AI
- Imagine How To Change: Explicit Procedure Modeling for Change Captioning
- Learning Brain Representation with Hierarchical Visual Embeddings
- Leveraging Data to Say No: Memory Augmented Plug-and-Play Selective Prediction
and 32 more in the interactive map.
CLIP and zero-shot · 39 papers
- MaskInversion: Localized Embeddings via Optimization of Explainability Maps (1 citation)
- Adaptive Concept Discovery for Interpretable Few-Shot Text Classification
- Automatic Image-Level Morphological Trait Annotation for Organismal Images
- Boosting Medical Visual Understanding From Multi-Granular Language Learning
- CLIP Behaves like a Bag-of-Words Model Cross-modally but not Uni-modally
- DVLA-RL: Dual-Level Vision–Language Alignment with Reinforcement Learning Gating for Few-Shot Learning
- Efficient Discriminative Joint Encoders for Large Scale Vision-Language Reranking
- Efficient Test-Time Scaling for Small Vision-Language Models
- Enhanced Continual Learning of Vision-Language Models with Model Fusion
- Hystar: Hypernetwork-driven Style-adaptive Retrieval via Dynamic SVD Modulation
- Let's Split Up: Zero-Shot Classifier Edits for Fine-Grained Video Understanding
- MergeTune: Continued Fine-Tuning of Vision-Language Models
and 27 more in the interactive map.
Visual question answering · 14 papers
- AQuA: Toward Strategic Response Generation for Ambiguous Visual Questions
- Choices Speak Louder than Questions
- EgoNight: Towards Egocentric Vision Understanding at Night with a Challenging Benchmark
- Query-Guided Spatial–Temporal–Frequency Interaction for Music Audio–Visual Question Answering
- RAVENEA: A Benchmark for Multimodal Retrieval-Augmented Visual Culture Understanding
- Wiki-R1: Incentivizing Multimodal Reasoning for Knowledge-based VQA via Data and Sampling Curriculum
- A Structured, Tagged, and Localized Visual Question Answering Dataset with Full Sentence Answers and Scene Graphs for Chest X-ray Images
- Adaptive Logit Adjustment for Debiasing Multimodal Language Models
- Can Vision-Language Models Answer Face to Face Questions in the Real-World?
- Compositional Neuro-Symbolic Concepts in Neural Activities
- Human Uncertainty-Aware Data Selection and Automatic Labeling in Visual Question Answering
- Knowledge Exchange with Confidence: Cost-Effective LLM Integration for Reliable and Efficient Visual Question Answering
and 2 more in the interactive map.