ICLR 2026PastMultimodal
ICLR 2026 Workshop on Multimodal Intelligence: Next Token Prediction & Beyond
ICLR 2026 Workshop MM Intelligence
- Submission deadline
- Feb 6, 2026, 13:00 UTCOpenReview-synced 2026-02-06 13:00 UTC (as of 2026-06-23) — extensions on OpenReview are applied automatically; verify on the website.
- Submission portal
- OpenReview
- Notes
- Topics were auto-suggested and may be imprecise — edits welcome.
Accepted papers (60)
Fetched from OpenReview (v2) on 2026-06-10.
A Single Image and Multimodality Is All You Need for Novel View Synthesis
A Systematic Study of Behavioral Cloning for Scientific Data Annotation
AdaTS: Adaptive Token Sampling for Efficient Speech Language Models
An Efficient Training Pipeline for Reasoning Graphical User Interface Agents
Beyond Seeing: Evaluating Multimodal LLMs on Tool-Enabled Image Perception, Transformation, and Reasoning
BlockGen: Flexible Blockwise Sequence Modeling with Hybrid Samplers
Bridging Generative and Predictive Paradigms via Hidden-Self-Distillation
Can Multi-Modal LLMs Provide Live Step-by-Step Task Guidance?
Can Vision Models Process Physiological Signals? Exploring Visual Tokenization as a Representation Interface
CARPE: Context-Aware Image Representation Prioritization via Ensemble for Large Vision-Language Models
City Navigation in the Wild: Exploring Emergent Navigation from Web-Scale Knowledge in MLLMs
CMRAG: Co-modality-based visual document retrieval and question answering
CRAFT-GUI: Curriculum-Reinforced Agent For GUI Tasks
Data Provenance for Image Auto-Regressive Generation
DeFacto: Counterfactual Thinking with Images for Enforcing Evidence-Grounded and Faithful Reasoning
Depth Over Specialization in Small Multimodal Transformers
Diagnosing the Curse: A Scale-Consistent and All-Phase Metric for Modality Bias in MLLMs
DiffuMamba: High-Throughput Diffusion LMs with Mamba Backbone
DISCO: Document Intelligence Suite for COmparative Evaluation
Does Semantic Noise Initialization Transfer from Images to Videos? A Paired Diagnostic Study
Efficient Multimodal Generation via Redundancy-Aware Mixture-of-Experts
ENACT: Evaluating Embodied Cognition with World Modeling of Egocentric Interaction
Esoteric Language Models: Bridging Autoregressive and Masked Diffusion LLMs
Fine-Tuning Masked Diffusion for Provable Self-Correction
GHVL: Geometry-Grounded Hyperbolic Vision-Language Models for Hierarchical Multimodal Representation Learning
GRAID: Enhancing Spatial Reasoning of VLMs through High-Fidelity Data Generation
Growing Visual Generative Capacity for Pre-Trained MLLMs
INDEX-PRESERVING LIGHTWEIGHT TOKEN PRUNING FOR EFFICIENT DOCUMENT UNDERSTANDING IN VISION-LANGUAGE MODELS
Infinity and Beyond: Compositional Alignment in VAR and Diffusion T2I Models
Is Extending Modality The Right Path Towards Omni-Modality?
LanteRn: Latent Visual Structured Reasoning
MapQA: A Map-Question-Answering Benchmark for Visual Language Model Reasoning
MLLMs are Deeply Affected by Modality Bias
Movie Facts and Fibs (MF$^2$): A Benchmark for Long Movie Understanding
Multimodal Language Models Cannot Spot Spatial Inconsistencies
Neural Signals Generate Clinical Notes in the Wild
Next-Scale Autoregression on Spectrograms for Sound Generation
Not All Features Are Created Equal: A Mechanistic Study of Vision-Language-Action Models
Reinforce Your Layout: Online Reward-Guided Diffusion for Layout-to-lmage Generation
Rethinking Visual Information Processing in Multimodal LLMs
RigidBench: Evaluating Rigid-Body Physics in Video Generation Models
Scaling Next-Brain-Token Prediction for MEG
SCOPE: Selective Cross-modal Orchestration of Visual Perception Experts
Seeing to Generalize: How Visual Data Corrects Binding Shortcuts
StarFlow: Generating Structured Workflow Outputs From Sketch Images
Stop Training for the Worst: Progressive Unmasking Accelerates Masked Diffusion Training
The Efficiency Gap in Byte Modeling
Towards Mitigating Hallucinations in Large Vision-Language Models by Refining Textual Embeddings
TowerVision: Understanding and Improving Multilinguality in Vision-Language Models
UniFusion: Vision-Language Model as Unified Encoder in Image Generation and Editing
Unifying Autoregressive and Discrete Diffusion Language Modeling via Cross-Regressive Decoding
Vid2Sid: Videos Can Help Close the Sim2Real Gap
VisGym: Diverse, Customizable, Scalable Environments for Multimodal Agents
Visual Representation Alignment for Multimodal Large Language Models
Visuals Lie, Consistency Speaks: Disentangling Spatial Attention from Reliability in Vision-Language Models
VLM-RobustBench: A Robustness Benchmark for Vision-Language Models
Worse Together: Understanding the Brittleness of Multimodal Models on Rare Concept Pairs
You Can Learn Tokenization End-to-End with Reinforcement Learning
Your Autoregressive Visual Model is a Natively Multi-Token Predictor : Speculative Coupled Decoding for Fast Autoregressive Visual Generation
Zoom-Zero: Reinforced Coarse-to-Fine Video Understanding via Temporal Zoom-in