InternVL3: Exploring Advanced Training and Test-Time Recipes for Open-Source Multimodal Models

2025/04/14 by Jinguo Zhu, Weiyun Wang, Zhu, Jinguo +95 · 1,656 citations
Computer Science · #Computer Vision and Pattern Recognition (cs.CV) #Domain Adaptation and Few-Shot Learning #FOS: Computer and information sciences #Multimodal Machine Learning Applications #Topic Modeling

paper · pdf · doi:10.48550/arxiv.2504.10479

openalex publication_date 2025/04/14 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

We introduce InternVL3, a significant advancement in the InternVL series featuring a native multimodal pre-training paradigm. Rather than adapting a text-only large language model (LLM) into a multimodal large language model (MLLM) that supports visual inputs, InternVL3 jointly acquires multimodal and linguistic capabilities from both diverse multimodal data and pure-text corpora during a single pre-training stage. This unified training paradigm effectively addresses the complexities and alignment challenges commonly encountered in conventional post-hoc training pipelines for MLLMs. To further improve performance and scalability, InternVL3 incorporates variable visual position encoding (V2PE) to support extended multimodal contexts, employs advanced post-training techniques such as supervised fine-tuning (SFT) and mixed preference optimization (MPO), and adopts test-time scaling strategies alongside an optimized training infrastructure. Extensive empirical evaluations demonstrate that InternVL3 delivers superior performance across a wide range of multi-modal tasks. In particular, InternVL3-78B achieves a score of 72.2 on the MMMU benchmark, setting a new state-of-the-art among open-source MLLMs. Its capabilities remain highly competitive with leading proprietary models, including ChatGPT-4o, Claude 3.5 Sonnet, and Gemini 2.5 Pro, while also maintaining strong pure-language proficiency. In pursuit of open-science principles, we will publicly release both the training data and model weights to foster further research and development in next-generation MLLMs.

Citations

• SmolVLM: Redefining small and efficient multimodal models
• VisualPRM: An Effective Process Reward Model for Multimodal Reasoning
• Qwen2.5-VL Technical Report
• SCP-116K: A High-Quality Problem-Solution Dataset and a Generalized Pipeline for Automated Extraction in the Higher Education Science Domain
• UI-TARS: Pioneering Automated GUI Interaction with Native Agents
• Eagle 2: Building Post-Training Data Strategies from Scratch for Frontier Vision-Language Models
• The Lessons of Developing Process Reward Models in Mathematical Reasoning
• AceMath: Advancing Frontier Math Reasoning with Post-Training and Reward Modeling
• Thinking in Space: How Multimodal Large Language Models See, Remember, and Recall Spaces
• CG-Bench: Clue-grounded Question Answering Benchmark for Long Video Understanding
• V2PE: Improving Multimodal Long-Context Capability of Vision-Language Models with Variable Visual Position Encoding
• Expanding Performance Boundaries of Open-Source Multimodal Models with Model, Data, and Test-Time Scaling
• Aguvis: Unified Pure Vision Agents for Autonomous GUI Interaction
• BlueLM-V-3B: Algorithm and System Co-Design for Multimodal Large Language Models on Mobile Devices
• Enhancing the Reasoning Ability of Multimodal Large Language Models via Mixed Preference Optimization
• OS-ATLAS: A Foundation Action Model for Generalist GUI Agents
• DynaMath: A Dynamic Visual Benchmark for Evaluating Mathematical Reasoning Robustness of Vision Language Models
• GPT-4o System Card
• Infinity-MM: Scaling Multimodal Performance with Large-Scale and High-Quality Instruction Data
• Mini-InternVL: A Flexible-Transfer Pocket Multimodal Model with 5% Parameters and 90% Performance
• R-Bench: Are your Large Multimodal Model Robust to Real-world Corruptions?
• TextHawk2: A Large Vision-Language Model Excels in Bilingual OCR and Grounding with 16x Fewer Tokens
• MM1.5: Methods, Analysis & Insights from Multimodal LLM Fine-tuning
• Oryx MLLM: On-Demand Spatial-Temporal Understanding at Arbitrary Resolution
• Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolution
• NVLM: Open Frontier-Class Multimodal LLMs
• Eagle: Exploring The Design Space for Multimodal LLMs with Mixture of Encoders
• MME-RealWorld: Could Your Multimodal LLM Challenge High-Resolution Real-World Scenarios that are Difficult for Humans?
• LLaVA-OneVision: Easy Visual Task Transfer
• Scaling LLM Test-Time Compute Optimally can be More Effective than Scaling Model Parameters
• MMIU: Multimodal Multi-image Understanding for Evaluating Large Vision-Language Models
• MiniCPM-V: A GPT-4V Level MLLM on Your Phone
• MM-Vet v2: A Challenging Benchmark to Evaluate Large Multimodal Models for Integrated Capabilities
• The Llama 3 Herd of Models
• LongVideoBench: A Benchmark for Long-context Interleaved Video-Language Understanding
• VLMEvalKit: An Open-Source Toolkit for Evaluating Large Multi-Modality Models
• LogicVista: Multimodal LLM Logical Reasoning Benchmark in Visual Contexts
• We-Math: Does Your Large Multimodal Model Achieve Human-like Mathematical Reasoning?
• LLM Critics Help Catch LLM Bugs
• CharXiv: Charting Gaps in Realistic Chart Understanding in Multimodal LLMs
• Math-LLaVA: Bootstrapping Mathematical Reasoning for Multimodal Large Language Models
• Long Context Transfer from Language to Vision
• Cambrian-1: A Fully Open, Vision-Centric Exploration of Multimodal LLMs
• MMBench-Video: A Long-Form Multi-Shot Benchmark for Holistic Video Understanding
• Benchmarking Multi-Image Understanding in Vision and Language Models: Perception, Knowledge, Reasoning, and Multi-Hop Reasoning
• WildVision: Evaluating Vision-Language Models in the Wild with Human Preferences
• MuirBench: A Comprehensive Benchmark for Robust Multi-image Understanding
• VideoLLaMA 2: Advancing Spatial-Temporal Modeling and Audio Understanding in Video-LLMs
• Improve Mathematical Reasoning in Language Models by Automated Process Supervision
• Parrot: Multilingual Visual Instruction Tuning
• Video-MME: The First-Ever Comprehensive Evaluation Benchmark of Multi-modal LLMs in Video Analysis
• Ovis: Structural Embedding Alignment for Multimodal Large Language Model
• M³CoT: A Novel Benchmark for Multi-Domain Multi-step Multi-modal Chain-of-Thought
• MTVQA: Benchmarking Multilingual Text-Centric Visual Question Answering
• How Far Are We to GPT-4V? Closing the Gap to Commercial Multimodal Models with Open-Source Suites
• SEED-Bench-2-Plus: Benchmarking Multimodal Large Language Models with Text-Rich Visual Comprehension
• MMT-Bench: A Comprehensive Multimodal Benchmark for Evaluating Large Vision-Language Models Towards Multitask AGI
• BLINK: Multimodal Large Language Models Can See but Not Perceive
• Ferret-v2: An Improved Baseline for Referring and Grounding with Large Language Models
• InternLM-XComposer2-4KHD: A Pioneering Large Vision-Language Model Handling Resolutions from 336 Pixels to 4K HD
• Binary Classifier Optimization for Large Language Model Alignment
• Are We on the Right Way for Evaluating Large Vision-Language Models?
• MathVerse: Does Your Multi-modal LLM Truly See the Diagrams in Visual Math Problems?
• Gemini 1.5: Unlocking multimodal understanding across millions of tokens of context
• The All-Seeing Project V2: Towards General Relation Comprehension of the Open World
• Measuring Multimodal Mathematical Reasoning with MATH-Vision Dataset
• InternEvo: Efficient Long-sequence Large Language Model Training via Hybrid Parallelism and Redundant Sharding
• SeeClick: Harnessing GUI Grounding for Advanced Visual GUI Agents
• InternVL: Scaling up Vision Foundation Models and Aligning for Generic Visual-Linguistic Tasks
• GeomVerse: A Systematic Evaluation of Large Models for Geometric Reasoning
• G-LLaVA: Solving Geometric Problem with Multi-Modal Large Language Model
• Math-Shepherd: Verify and Reinforce LLMs Step-by-step without Human Annotations
• VILA: On Pre-training for Visual Language Models
• MVBench: A Comprehensive Multi-modal Video Understanding Benchmark
• MMMU: A Massive Multi-discipline Multimodal Understanding and Reasoning Benchmark for Expert AGI
• Monkey: Image Resolution and Text Label Are Important Things for Large Multi-modal Models
• mPLUG-Owl2: Revolutionizing Multi-modal Large Language Model with Modality Collaboration
• CogVLM: Visual Expert for Pretrained Language Models
• HallusionBench: An Advanced Diagnostic Suite for Entangled Language Hallucination and Visual Illusion in Large Vision-Language Models
• MathVista: Evaluating Mathematical Reasoning of Foundation Models in Visual Contexts
• Aligning Large Multimodal Models with Factually Augmented RLHF
• MM-Vet: Evaluating Large Multimodal Models for Integrated Capabilities
• The All-Seeing Project: Towards Panoptic Visual Recognition and Understanding of the Open World
• MMBench: Is Your Multi-modal Model an All-around Player?
• Shikra: Unleashing Multimodal LLM's Referential Dialogue Magic
• MME: A Comprehensive Evaluation Benchmark for Multimodal Large Language Models
• CMMLU: Measuring massive multitask language understanding in Chinese
• Let's Verify Step by Step
• Direct Preference Optimization: Your Language Model is Secretly a Reward Model
• Evaluating the Performance of Large Language Models on GAOKAO Benchmark
• TheoremQA: A Theorem-driven Question Answering dataset
• ONE-PEACE: Exploring One General Representation Model Toward Unlimited Modalities
• Evaluating Object Hallucination in Large Vision-Language Models
• C-Eval: A Multi-Level Multi-Discipline Chinese Evaluation Suite for Foundation Models
• VideoChat: Chat-Centric Video Understanding
• Visual Instruction Tuning
• GPT-4 Technical Report
• Universal Instance Perception as Object Discovery and Retrieval
• Annotated Point Clouds and Images from a Spatial-Semantic Perception Pipeline: Robotized Deconstruction at the Reference Construction Site, Aachen
• MapQA: A Dataset for Question Answering on Choropleth Maps
• Challenging BIG-Bench Tasks and Whether Chain-of-Thought Can Solve Them
• Learn to Explain: Multimodal Reasoning via Thought Chains for Science Question Answering
• CLEVR-Math: A Dataset for Compositional Language, Visual and Mathematical Reasoning
• ChartQA: A Benchmark for Question Answering about Charts with Visual and Logical Reasoning
• MViTv2: Improved Multiscale Vision Transformers for Classification and Detection
• Training Verifiers to Solve Math Word Problems
• IconQA: A New Benchmark for Abstract Diagram Understanding and Visual Language Reasoning
• Program Synthesis with Large Language Models
• Evaluating Large Language Models Trained on Code
• Inter-GPS: Interpretable Geometry Problem Solving with Formal Language and Symbolic Reasoning
• InfographicVQA
• Measuring Mathematical Problem Solving With the MATH Dataset
• Measuring Massive Multitask Language Understanding
• DocVQA: A Dataset for VQA on Document Images
• ICDAR2019 Competition on Scanned Receipt OCR and Information Extraction
• Scene Text Visual Question Answering
• OK-VQA: A Visual Question Answering Benchmark Requiring External\n Knowledge
• HellaSwag: Can a Machine Really Finish Your Sentence?
• Investigating Prior Knowledge for Challenging Chinese Machine Reading Comprehension
• Towards VQA Models That Can Read
• DVQA: Understanding Data Visualizations via Question Answering
• Simple and Effective Multi-Paragraph Reading Comprehension
• TriviaQA: A Large Scale Distantly Supervised Challenge Dataset for\n Reading Comprehension
• RACE: Large-scale ReAding Comprehension Dataset From Examinations
• Making the V in VQA Matter: Elevating the Role of Image Understanding in Visual Question Answering
• A Diagram Is Worth A Dozen Images
• Generation and Comprehension of Unambiguous Object Descriptions
• Molmo and PixMo: Open Weights and Open Data for State-of-the-Art Vision-Language Models
• VCR: A Task for Pixel-Level Complex Reasoning in Vision Language Models via Restoring Occluded Text
• MLVU: Benchmarking Multi-task Long Video Understanding
• Qwen-VL: A Versatile Vision-Language Model for Understanding, Localization, Text Reading, and Beyond

Cited by

• MIRA: Multimodal Iterative Reasoning Agent for Image Editing
• From Generated Human Videos to Physically Plausible Robot Trajectories
• ReasonX: MLLM-Guided Intrinsic Image Decomposition
• CME-CAD: Heterogeneous Collaborative Multi-Expert Reinforcement Learning for CAD Code Generation
• MM-UAVBench: How Well Do Multimodal Large Language Models See, Think, and Plan in Low-Altitude UAV Scenarios?
• Video-BrowseComp: Benchmarking Agentic Video Research on Open Web
• MUSON: A Reasoning-oriented Multimodal Dataset for Socially Compliant Navigation in Urban Environments
• Modality Inflation: Energy Characterization and Optimization Opportunities for MLLM Inference
• Self-Rewarded Multimodal Coherent Reasoning Across Diverse Visual Domains
• Dream-VL & Dream-VLA: Open Vision-Language and Vision-Language-Action Models with Diffusion Language Model Backbone
• FETAL-GAUGE: A Benchmark for Assessing Vision-Language Models in Fetal Ultrasound
• LiveProteinBench: A Contamination-Free Benchmark for Assessing Models' Specialized Capabilities in Protein Science
• Masking Teacher and Reinforcing Student for Distilling Vision-Language Models
• VideoScaffold: Elastic-Scale Visual Hierarchies for Streaming Video Understanding in MLLMs
• SoliReward: Mitigating Susceptibility to Reward Hacking and Annotation Noise in Video Generation Reward Models
• MarineEVT: Advancing Event-Centric Marine Video Understanding via Visual Tool Reasoning
• ObsDriveBench: Benchmarking Multimodal Understanding under Adverse Weather with Observability Awareness
• Offline-Online Curriculum RL for Multimodal Reasoning
• UltraViT: Latency-Optimized On-device Vision Encoder for Large Vision-Language Models
• ODYSSE: Episode-wise Policy Optimization for Personalized Agentic Reasoning
• Mage-VL: An Efficient Codec-Native Streaming Multimodal Foundation Model
• What You See Is What You Get: Observation-Aligned Supervision for Chart-to-Code Generation
• See Less, See Right: Bi-directional Perceptual Shaping For Multimodal Reasoning
• An LMM for Precisely Grounding Elements in Documents
• RoboProcessBench: Benchmarking Process-Aware Understanding in Vision-Language Robotic Manipulation
• VLD-RAG: Agentic Vision-Language Retrieval-Augmented Generation for Long, Visually-Rich Multi-Page Documents
• Perceive and Calibrate: Analyzing and Enhancing Robustness of Medical Multi-Modal Large Language Models
• EraseLoRA: MLLM-Driven Foreground Exclusion and Background Subtype Aggregation for Dataset-Free Object Removal
• UniPercept: Towards Unified Perceptual-Level Image Understanding across Aesthetics, Quality, Structure, and Texture
• Streaming Video Instruction Tuning
• MarineEval: Assessing the Marine Intelligence of Vision-Language Models
• UniRec-0.1B: Unified Text and Formula Recognition with 0.1B Parameters
• MMSRARec: Summarization and Retrieval Augumented Sequential Recommendation Based on Multimodal Large Language Model
• Benchmarking and Enhancing VLM for Compressed Image Understanding
• VL4Gaze: Unleashing Vision-Language Models for Gaze Following
• Beyond Motion Pattern: An Empirical Study of Physical Forces for Human Motion Understanding
• Generative Human-Object Interaction Detection via Differentiable Cognitive Steering of Multi-modal LLMs
• Bridging Semantics and Geometry: A Decoupled LVLM-SAM Framework for Reasoning Segmentation in Optical Remote Sensing
• IPCV: Information-Preserving Compression for MLLM Visual Encoders
• Thinking Beyond Labels: Vocabulary-Free Fine-Grained Recognition using Reasoning-Augmented LMMs
• FPBench: A Comprehensive Benchmark of Multimodal Large Language Models for Fingerprint Analysis
• Enabling Disaggregated Multi-Stage MLLM Inference via GPU-Internal Scheduling and Resource Sharing
• Xiaomi MiMo-VL-Miloco Technical Report
• Deep But Reliable: Advancing Multi-turn Reasoning for Thinking with Images
• AdaTooler-V: Adaptive Tool-Use for Images and Videos
• A Benchmark and Agentic Framework for Omni-Modal Reasoning and Tool Use in Long Videos
• AMUSE: Audio-Visual Benchmark and Alignment Framework for Agentic Multi-Speaker Understanding
• Scaling Spatial Reasoning in MLLMs through Programmatic Data Synthesis
• Vision-Language-Action Models for Autonomous Driving: Past, Present, and Future
• Seeing Beyond the Scene: Analyzing and Mitigating Background Bias in Action Recognition
• Uni-Parser Technical Report
• V-REX: Benchmarking Exploratory Visual Reasoning via Chain-of-Questions
• JMMMU-Pro: Image-based Japanese Multi-discipline Multimodal Understanding Benchmark via Vibe Benchmark Construction
• WorldPlay: Towards Long-Term Geometric Consistency for Real-Time Interactive World Modeling
• Enhancing Visual Programming for Visual Reasoning via Probabilistic Graphs
• Ophiuchus: Incentivizing Tool-augmented "Think with Images" for Joint Medical Segmentation, Understanding and Reasoning
• Cornfigurator: Automated Planning for Any-to-Any Multimodal Model Serving
• HyperVL: An Efficient and Dynamic Multimodal Large Language Model for Edge Devices
• KFS-Bench: Comprehensive Evaluation of Key Frame Sampling in Long Video Understanding
• MobileWorldBench: Towards Semantic World Modeling For Mobile Agents
• A4-Agent: An Agentic Framework for Zero-Shot Affordance Reasoning
• SAGE: Training Smart Any-Horizon Agents for Long Video Reasoning with Reinforcement Learning
• Adapting MLLMs for Nuanced Video Retrieval
• Spatial-Aware VLA Pretraining through Visual-Physical Alignment from Human Videos
• DrivePI: Spatial-aware 4D MLLM for Unified Autonomous Driving Understanding, Perception, Prediction and Planning
• FysicsWorld: A Unified Full-Modality Benchmark for Any-to-Any Understanding, Generation, and Reasoning
• CogDoc: Towards Unified thinking in Documents
• More Than the Final Answer: Improving Visual Extraction and Logical Consistency in Vision-Language Models
• Rethinking Jailbreak Detection of Large Vision Language Models with Representational Contrastive Scoring
• Exploring MLLM-Diffusion Information Transfer with MetaCanvas
• SmokeBench: Evaluating Multimodal Large Language Models for Wildfire Smoke Detection
• HFS: Holistic Query-Aware Frame Selection for Efficient Video Reasoning
• VGent: Visual Grounding via Modular Design for Disentangling Reasoning and Prediction
• MMSI-Video-Bench: A Holistic Benchmark for Video-Based Spatial Intelligence
• Tool-Augmented Spatiotemporal Reasoning for Streamlining Video Question Answering Task
• Investigating The Functional Roles of Attention Heads in Vision Language Models: Evidence for Reasoning Modules
• Grounding Everything in Tokens for Multimodal Large Language Models
• DynaIP: Dynamic Image Prompt Adapter for Scalable Zero-shot Personalized Text-to-Image Generation
• Rethinking Chain-of-Thought Reasoning for Videos
• GimbalDiffusion: Gravity-Aware Camera Control for Video Generation
• ARM-Thinker: Reinforcing Multimodal Generative Reward Models with Agentic Tool Use and Visual Reasoning
• Same Content, Different Answers: Cross-Modal Inconsistency in MLLMs
• From Segments to Scenes: Temporal Understanding for Agentic Autonomous Driving via Vision-Language Models
• SpatialDreamer: Incentivizing Spatial Reasoning via Active Mental Imagery
• Geo3DVQA: Evaluating Vision-Language Models for 3D Geospatial Reasoning from Aerial Imagery
• ParaUni: Enhance Generation in Unified Multimodal Model with Reinforcement-driven Hierarchical Parallel Information Interaction
• VisChainBench: A Benchmark for Multi-Turn, Multi-Image Visual Reasoning Beyond Language Priors
• ProAgent: Harnessing On-Demand Sensory Contexts for Proactive LLM Agent Systems
• The Role of Entropy in Visual Grounding: Analysis and Optimization
• RefBench-PRO: Perceptual and Reasoning Oriented Benchmark for Referring Expression Comprehension
• WAM-Flow: Parallel Coarse-to-Fine Motion Planning via Discrete Flow Matching for Autonomous Driving
• The Dynamic Prior: Understanding 3D Structures for Casual Dynamic Videos
• MemLoRA: Distilling Expert Adapters for On-Device Memory Systems
• StreamEQA: Towards Streaming Video Understanding for Embodied Scenarios
• Jina-VLM: Small Multilingual Vision Language Model
• Colon-X: Advancing Intelligent Colonoscopy toward Clinical Reasoning
• ToG-Bench: Task-Oriented Spatio-Temporal Grounding in Egocentric Videos
• Text-Printed Image: Bridging the Image-Text Modality Gap for Text-centric Training of Large Vision-Language Models
• Divide, then Ground: Adapting Frame Selection to Query Types for Long-Form Video Understanding
• LoVoRA: Text-guided and Mask-free Video Object Removal and Addition with Learnable Object-aware Localization
• MRD: Multi-resolution Retrieval-Detection Fusion for High-Resolution Image Understanding
• MindGPT-4ov: An Enhanced MLLM via a Multi-Stage Post-Training Paradigm
• Diagnose, Correct, and Learn from Manipulation Failures via Visual Symbols
• VACoT: Rethinking Visual Data Augmentation with VLMs
• Beyond N-grams: A Hierarchical Reward Learning Framework for Clinically-Aware Medical Report Generation
• Script: Graph-Structured and Query-Conditioned Semantic Token Pruning for Multimodal Large Language Models
• Guardian: Detecting Robotic Planning and Execution Errors with Vision-Language Models
• ViRectify: A Challenging Benchmark for Video Reasoning Correction with Multimodal Large Language Models
• CycliST: A Video Language Model Benchmark for Reasoning on Cyclical State Transitions
• ChartAnchor: Chart Grounding with Structural-Semantic Fidelity
• PhotoFramer: Multi-modal Image Composition Instruction
• Thinking with Drafts: Speculative Temporal Reasoning for Efficient Long Video Understanding
• Better, Stronger, Faster: Tackling the Trilemma in MLLM-based Segmentation with Simultaneous Textual Mask Prediction
• Video-R2: Reinforcing Consistent and Grounded Reasoning in Multimodal Language Models
• Video-CoM: Interactive Video Reasoning via Chain of Manipulations
• Visual Generation Tuning
• VQRAE: Representation Quantization Autoencoders for Multimodal Understanding, Generation and Reconstruction
• MathSight: A Benchmark Exploring Have Vision-Language Models Really Seen in University-Level Mathematical Reasoning?
• SpaceMind: Camera-Guided Modality Fusion for Spatial Reasoning in Vision-Language Models
• AREA3D: Active Reconstruction Agent with Unified Feed-Forward 3D Perception and Vision-Language Guidance
• UniGeoSeg: Towards Unified Open-World Segmentation for Geospatial Scenes
• World in a Frame: Understanding Culture Mixing as a New Challenge for Vision-Language Models
• AIA: Rethinking Architecture Decoupling Strategy In Unified Multimodal Model
• RoadSceneBench: A Lightweight Benchmark for Mid-Level Road Scene Understanding
• Agentic Learner with Grow-and-Refine Multimodal Semantic Memory
• Multi-Crit: Benchmarking Multimodal Judges on Pluralistic Criteria-Following
• Guiding the Inner Eye: A Framework for Hierarchical and Flexible Visual Grounded Reasoning
• Thinking With Bounding Boxes: Enhancing Spatio-Temporal Video Grounding via Reinforcement Fine-Tuning
• Co-Training Vision Language Models for Remote Sensing Multi-task Learning
• LungNoduleAgent: A Collaborative Multi-Agent System for Precision Diagnosis of Lung Nodules
• CaptionQA: Is Your Caption as Useful as the Image Itself?
• Text-Guided Semantic Image Encoder
• SPHINX: A Synthetic Environment for Visual Perception and Reasoning
• NVIDIA Nemotron Parse 1.1
• MTBBench: A Multimodal Sequential Clinical Decision-Making Benchmark in Oncology
• WaymoQA: A Multi-View Visual Question Answering Dataset for Safety-Critical Reasoning in Autonomous Driving
• Semantic Router: On the Feasibility of Hijacking MLLMs via a Single Adversarial Perturbation
• Boosting Reasoning in Large Multimodal Models via Activation Replay
• Reasoning-VLA: A Fast and General Vision-Language-Action Reasoning Model for Autonomous Driving
• MonoSR: Open-Vocabulary Spatial Reasoning from Monocular Images
• Temporal-Visual Semantic Alignment: A Unified Architecture for Transferring Spatial Priors from Vision Models to Zero-Shot Temporal Tasks
• Scaling Agentic Reinforcement Learning for Tool-Integrated Reasoning in VLMs
• Be My Eyes: Extending Large Language Models to New Modalities Through Multi-Agent Collaboration
• LAST: LeArning to Think in Space and Time for Generalist Vision-Language Models
• VideoChat-M1: Collaborative Policy Planning for Video Understanding via Multi-Agent Reinforcement Learning
• Table Comprehension in Building Codes using Vision Language Models and Domain-Specific Fine-Tuning
• DuoTeach: Dual Role Self-Teaching for Coarse-to-Fine Decision Coordination in Vision--Language Models
• VideoPerceiver: Enhancing Fine-Grained Temporal Perception in Video Multimodal Large Language Models
• Beyond Description: Cognitively Benchmarking Fine-Grained Action for Embodied Agents
• MedVision: Benchmarking Quantitative Medical Image Analysis
• Chain-of-Visual-Thought: Teaching VLMs to See and Think Better with Continuous Visual Tokens
• Disc3D: Automatic Curation of High-Quality 3D Dialog Data via Discriminative Object Referring
• Mixture of Horizons in Action Chunking
• DocPTBench: Benchmarking End-to-End Photographed Document Parsing and Translation
• ChineseVideoBench: Benchmarking Multi-modal Large Models for Chinese Video Question Answering
• MASS: Motion-Aware Spatial-Temporal Grounding for Physics Reasoning and Comprehension in Vision-Language Models
• ConsistCompose: Unified Multimodal Layout Control for Image Composition
• Can a Second-View Image Be a Language? Geometric and Semantic Cross-Modal Reasoning for X-ray Prohibited Item Detection
• Multi-speaker Attention Alignment for Multimodal Social Interaction
• ChemVTS-Bench: Evaluating Visual-Textual-Symbolic Reasoning of Multimodal Large Language Models in Chemistry
• VITAL: Vision-Encoder-centered Pre-training for LMMs in Visual Quality Assessment
• When Better Teachers Don't Make Better Students: Revisiting Knowledge Distillation for CLIP Models in VQA
• The Potential and Limitations of Vision-Language Models for Human Motion Understanding: A Case Study in Data-Driven Stroke Rehabilitation
• Q-REAL: Towards Realism and Plausibility Evaluation for AI-Generated Content
• Lost in Translation and Noise: A Deep Dive into the Failure Modes of VLMs on Real-World Tables
• EvoLMM: Self-Evolving Large Multimodal Models with Continuous Rewards
• Learning to Think Fast and Slow for Visual Language Models
• TeamPath: Building MultiModal Pathology Experts with Reasoning AI Copilots
• POMA-3D: The Point Map Way to 3D Scene Understanding
• Text2Traffic: A Text-to-Image Generation and Editing Method for Traffic Scenes
• OpenMMReasoner: Pushing the Frontiers for Multimodal Reasoning with an Open and General Recipe
• CrossCheck-Bench: Diagnosing Compositional Failures in Multimodal Conflict Resolution
• SciEGQA: A Dataset for Scientific Evidence-Grounded Question Answering and Reasoning
• MoDES: Accelerating Mixture-of-Experts Multimodal Large Language Models via Dynamic Expert Skipping
• Evaluating Multimodal Large Language Models on Vertically Written Japanese Text
• FarSLIP: Discovering Effective CLIP Adaptation for Fine-Grained Remote Sensing Understanding
• ManipShield: A Unified Framework for Image Manipulation Detection, Localization and Explanation
• P1: Mastering Physics Olympiads with Reinforcement Learning
• GeoX-Bench: Benchmarking Cross-View Geo-Localization and Pose Estimation Capabilities of Large Multimodal Models
• DualTAP: A Dual-Task Adversarial Protector for Mobile MLLM Agents
• Seeing isn't Hearing: Benchmarking Vision Language Models at Interpreting Spectrograms
• MOON2.0: Dynamic Modality-balanced Multimodal Representation Learning for E-commerce Product Understanding
• Imagine in Space: Exploring the Frontier of Spatial Intelligence and Reasoning Efficiency in Vision Language Models
• CrossVid: A Comprehensive Benchmark for Evaluating Cross-Video Reasoning in Multimodal Large Language Models
• KVSwap: Disk-aware KV Cache Offloading for Long-Context On-device Inference
• EcoAlign: An Economically Rational Framework for Efficient LVLM Alignment
• Enhancing the Outcome Reward-based RL Training of MLLMs with Self-Consistency Sampling
• Towards Trustworthy Dermatology MLLMs: A Benchmark and Multimodal Evaluator for Diagnostic Narratives
• MACEval: A Multi-Agent Continual Evaluation Network for Large Models
• mmJEE-Eval: A Bilingual Multimodal Benchmark for Evaluating Scientific Reasoning in Vision-Language Models
• UI2Code^N: A Visual Language Model for Test-Time Scalable Interactive UI-to-Code Generation
• Federated Learning for Video Violence Detection: Complementary Roles of Lightweight CNNs and Vision-Language Models for Energy-Efficient Use
• Improving Region Representation Learning from Urban Imagery with Noisy Long-Caption Supervision
• MVU-Eval: Towards Multi-Video Understanding Evaluation for Multimodal LLMs
• Culture in Action: Evaluating Text-to-Image Models through Social Activities
• DeepEyesV2: Toward Agentic Multimodal Model
• Visual Spatial Tuning
• Cambrian-S: Towards Spatial Supersensing in Video
• ThaiOCRBench: A Task-Diverse Benchmark for Vision-Language Understanding in Thai
• V-Thinker: Interactive Thinking with Images
• Surprisal reveals diversity gaps in image captioning and different scorers change the story
• Evo-1: Lightweight Vision-Language-Action Model with Preserved Semantic Alignment
• SurgViVQA: Temporally-Grounded Video Question Answering for Surgical Scene Understanding
• VCode: a Multimodal Coding Benchmark with SVG as Symbolic Visual Representation
• Can Visual Input Be Compressed? A Visual Token Compression Benchmark for Large Multimodal Models
• When Visualizing is the First Step to Reasoning: MIRA, a Benchmark for Visual Chain-of-Thought
• TIR-Bench: A Comprehensive Benchmark for Agentic Thinking-with-Images Reasoning
• Fleming-VL: Towards Universal Medical Visual Reasoning with Multimodal LLMs
• OmniBrainBench: A Comprehensive Multimodal Benchmark for Brain Imaging Analysis Across Multi-stage Clinical Tasks
• CueBench: Advancing Unified Understanding of Context-Aware Video Anomalies in Real-World
• Saliency-R1: Incentivizing Unified Saliency Reasoning Capability in MLLM with Confidence-Guided Reinforcement Learning
• VinciCoder: Unifying Multimodal Code Generation via Coarse-to-fine Visual Reinforcement Learning
• Rethinking Facial Expression Recognition in the Era of Multimodal Large Language Models: Benchmark, Datasets, and Beyond
• Foundation Models for Trajectory Planning in Autonomous Driving: A Review of Progress and Open Challenges
• PETAR: Localized Findings Generation with Mask-Aware Vision-Language Modeling for PET Automated Reporting
• FLoC: Facility Location-Based Efficient Visual Token Compression for Long Video Understanding
• ToolScope: An Agentic Framework for Vision-Guided and Long-Horizon Tool Use
• SlideAgent: Hierarchical Agentic Framework for Multi-Page Visual Document Understanding
• Counteracting Matthew Effect in Self-Improvement of LVLMs through Head-Tail Re-balancing
• LoCoT2V-Bench: A Benchmark for Long-Form and Complex Text-to-Video Generation
• PairUni: Pairwise Training for Unified Multimodal Language Models
• RecoReward: Recommender-Guided Multimodal Description Generation for Recommendation
• ChartNet: A Million-Scale, High-Quality Multimodal Dataset for Robust Chart Understanding
• TS-Reasoner: Aligning Time Series Foundation Models with LLM Reasoning
• InstructPLM-mu: 1-Hour Fine-Tuning of ESM2 Beats ESM3 in Protein Mutation Predictions
• Ouroboros-Spatial: Closing the Data-Model Loop for Spatial Reasoning
• Unlimited OCR Works
• PRISMM-Bench: A Benchmark of Peer-Review Grounded Multimodal Inconsistencies
• OSWorld-MCP: Benchmarking MCP Tool Invocation In Computer-Use Agents
• ViPER: Empowering the Self-Evolution of Visual Perception Abilities in Vision-Language Model
• Video-Thinker: Sparking "Thinking with Videos" via Reinforcement Learning
• Omni-Reward: Towards Generalist Omni-Modal Reward Modeling with Free-Form Preferences
• VideoTG-R1: Boosting Video Temporal Grounding via Curriculum Reinforcement Learning on Reflected Boundary Annotations
• DynaSolidGeo: A Dynamic Benchmark for Genuine Spatial Mathematical Reasoning of VLMs in Solid Geometry
• OmniVinci: Enhancing Architecture and Data for Omni-Modal Understanding LLM
• Towards Physically Executable 3D Gaussian for Embodied Navigation
• SafetyPairs: Isolating Safety Critical Image Features with Counterfactual Image Generation
• Small Drafts, Big Verdict: Information-Intensive Visual Reasoning via Speculation
• GeoThought: A Dataset for Enhancing Mathematical Geometry Reasoning in Vision-Language Models
• BioCAP: Exploiting Synthetic Captions Beyond Labels in Biological Foundation Models
• Open-o3-Video: Grounded Video Reasoning with Explicit Spatio-Temporal Evidence
• Mixing Importance with Diversity: Joint Optimization for KV Cache Compression in Large Vision-Language Models
• Metis-HOME: Hybrid Optimized Mixture-of-Experts for Multimodal Reasoning
• MedReason-R1: Learning to Reason for CT Diagnosis with Reinforcement Learning and Local Zoom
• Detecting Latin in Historical Books with Large Language Models: A Multimodal Benchmark
• [De|Re]constructing VLMs' Reasoning in Counting
• DaMo: Data Mixing Optimizer in Fine-tuning Multimodal LLMs for Mobile Phone Agents
• Unified Reinforcement and Imitation Learning for Vision-Language Models
• Seeing Across Views: Benchmarking Spatial Reasoning of Vision-Language Models in Robotic Scenes
• VITA-E: Natural Embodied Interaction with Concurrent Seeing, Hearing, Speaking, and Acting
• VDRive: Leveraging Reinforced VLA and Diffusion Policy for End-to-end Autonomous Driving
• When Seeing Is not Enough: Revealing the Limits of Active Reasoning in MLLMs
• Proactive Reasoning-with-Retrieval Framework for Medical Multimodal Large Language Models
• Grasp Any Region: Towards Precise, Contextual Pixel Understanding for Multimodal LLMs
• DeepSeek-OCR: Contexts Optical Compression
• World-in-World: World Models in a Closed-Loop World
• Think with 3D: Geometric Imagination Grounded Spatial Reasoning from Limited Views
• VERA-V: Variational Inference Framework for Jailbreaking Vision-Language Models
• iDETEX: Empowering MLLMs for Intelligent DETailed EXplainable IQA
• SimpleVSF: VLM-Scoring Fusion for Trajectory Prediction of End-to-End Autonomous Driving
• See or Say Graphs: Agent-Driven Scalable Graph Structure Understanding with Vision-Language Models
• GACO-CAD: Geometry-Augmented and Conciseness-Optimized CAD Model Generation from Single Image
• Infinity Parser: Layout Aware Reinforcement Learning for Scanned Document Parsing
• Training-free Online Video Step Grounding
• Pursuing Minimal Sufficiency in Spatial Reasoning
• Diagnosing Bottlenecks in Data Visualization Understanding by Vision-Language Models
• Select Less, Reason More: Prioritizing Evidence Purity for Video Reasoning
• From Pixels to Words -- Towards Native Vision-Language Primitives at Scale
• You May Speak Freely: Improving the Fine-Grained Visual Recognition Capabilities of Multimodal Large Language Models with Answer Extraction
• Benchmarking Multimodal Large Language Models for Face Recognition
• Adapting Self-Supervised Representations as a Latent Space for Efficient Generation
• Can MLLMs Absorb Math Reasoning Abilities from LLMs as Free Lunch?
• Knowledge-based Visual Question Answer with Multimodal Processing, Retrieval and Filtering
• InteractiveOmni: A Unified Omni-modal Model for Audio-Visual Multi-turn Dialogue
• Spatial-DISE: A Unified Benchmark for Evaluating Spatial Reasoning in Vision-Language Models
• InternVLA-M1: A Spatially Guided Vision-Language-Action Framework for Generalist Robot Policy
• Scope: Selective Cross-modal Orchestration of Visual Perception Experts
• Detect Anything via Next Point Prediction
• UniFusion: Vision-Language Model as Unified Encoder in Image Generation
• VideoLucy: Deep Memory Backtracking for Long Video Understanding
• SafeMT: Multi-turn Safety for Multimodal Language Models
• MetaCaptioner: Towards Generalist Visual Captioning with Open-source Suites
• Epistemic-aware Vision-Language Foundation Model for Fetal Ultrasound Interpretation
• ViCO: A Training Strategy towards Semantic Aware Dynamic High-Resolution
• HoneyBee: Data Recipes for Vision-Language Reasoners
• ExpVid: A Benchmark for Experiment Video Understanding & Reasoning
• ODI-Bench: Can MLLMs Understand Immersive Omnidirectional Environments?
• InternSVG: Towards Unified SVG Tasks with Multimodal Large Language Models
• BLEnD-Vis: Benchmarking Multimodal Cultural Understanding in Vision Language Models
• Vlaser: Vision-Language-Action Model with Synergistic Embodied Reasoning
• A Survey on Agentic Multimodal Large Language Models
• Catch-Only-One: Non-Transferable Examples for Model-Specific Authorization
• Judge Before Answer: Can MLLM Discern the False Premise in Question?
• FG-CLIP 2: A Bilingual Fine-grained Vision-Language Alignment Model
• RefineShot: Rethinking Cinematography Understanding with Foundational Skill Evaluation
• Image-to-Video Transfer Learning based on Image-Language Foundation Models: A Comprehensive Survey
• OmniQuality-R: Advancing Reward Models Through All-Encompassing Quality Assessment
• ViSurf: Visual Supervised-and-Reinforcement Fine-Tuning for Large Vision-and-Language Models
• UniFlow: A Unified Pixel Flow Tokenizer for Visual Understanding and Generation
• Taming a Retrieval Framework to Read Images in Humanlike Manner for Augmenting Generation of MLLMs
• Stacked Regression using Off-the-shelf, Stimulus-tuned and Fine-tuned Neural Networks for Predicting fMRI Brain Responses to Movies (Algonauts 2025 Report)
• OpusAnimation: Code-Based Dynamic Chart Generation
• Semantic Visual Anomaly Detection and Reasoning in AI-Generated Images
• Think Twice to See More: Iterative Visual Reasoning in Medical VLMs
• Patch-as-Decodable-Token: Towards Unified Multi-Modal Vision Tasks in MLLMs
• AutoPR: Let's Automate Your Academic Promotion!
• Look Less, Reason More: Rollout-Guided Adaptive Pixel-Space Reasoning
• Towards Efficient Multimodal Unified Reasoning Model via Model Merging
• Constructive Distortion: Improving MLLMs with Attention-Guided Image Warping
• Unleashing Perception-Time Scaling to Multimodal Reasoning Models
• PhysToolBench: Benchmarking Physical Tool Understanding for MLLMs
• Multimodal Prompt Optimization: Why Not Leverage Multiple Modalities for MLLMs
• Thinking with Camera: A Unified Multimodal Model for Camera-Centric Understanding and Generation
• To Sink or Not to Sink: Visual Information Pathways in Large Vision-Language Models
• Evaluating Small Vision-Language Models on Distance-Dependent Traffic Perception
• Kelp: A Streaming Safeguard for Large Models via Latent Dynamics-Guided Risk Detection
• MARC: Memory-Augmented RL Token Compression for Efficient Video Understanding
• GTR-Bench: Evaluating Geo-Temporal Reasoning in Vision-Language Models
• VideoNorms: Benchmarking Cultural Awareness of Video Language Models
• CIR-CoT: Towards Interpretable Composed Image Retrieval via End-to-End Chain-of-Thought Reasoning
• RetouchLLM: Training-free Code-based Image Retouching with Vision Language Models
• SciVideoBench: Benchmarking Scientific Video Reasoning in Large Multimodal Models
• Growing Visual Generative Capacity for Pre-Trained MLLMs
• TTRV: Test-Time Reinforcement Learning for Vision Language Models
• StaR-KVQA: Structured Reasoning Traces for Implicit-Knowledge Visual Question Answering
• Lumina-DiMOO: An Omni Diffusion Large Language Model for Multi-Modal Generation and Understanding
• Fine-grained Defocus Blur Control for Generative Image Models
• HOI-R1: Exploring the Potential of Multimodal Large Language Models for Human-Object Interaction Detection
• D2E: Scaling Vision-Action Pretraining on Desktop Data for Transfer to Embodied AI
• A.I.R.: Enabling Adaptive, Iterative, and Reasoning-based Frame Selection For Video Question Answering
• AgriGPT-VL: Agricultural Vision-Language Understanding Suite
• What Shapes a Creative Machine Mind? Comprehensively Benchmarking Creativity in Foundation Models
• Person-Centric Annotations of LAION-400M: Auditing Bias and Its Transfer to Models
• Exploring Instruction Data Quality for Explainable Image Quality Assessment
• Agentic Jigsaw Interaction Learning for Enhancing Visual Perception and Reasoning in Vision-Language Models
• Automated Structured Radiology Report Generation with Rich Clinical Context
• PAL-UI: Planning with Active Look-back for Vision-Based GUI Agents
• TAMA: Tool-Augmented Multimodal Agent for Procedural Activity Understanding
• Learning to See Before Seeing: Demystifying LLM Visual Priors from Language Pre-training
• Clarification as Supervision: Reinforcement Learning for Vision-Language Interfaces
• TimeScope: Towards Task-Oriented Temporal Grounding In Long Videos
• Human-MME: A Holistic Evaluation Benchmark for Human-Centric Multimodal Large Language Models
• Towards Continual Expansion of Data Coverage: Automatic Text-guided Edge-case Synthesis
• DeepSketcher: Internalizing Visual Manipulation for Multimodal Reasoning
• MuSLR: Multimodal Symbolic Logical Reasoning
• Interpret, prune and distill Donut : towards lightweight VLMs for VQA on document
• VLM-FO1: Bridging the Gap Between High-Level Reasoning and Fine-Grained Perception in VLMs
• LLaVAShield: Safeguarding Multimodal Multi-Turn Dialogues in Vision-Language Models
• Vision-Zero: Scalable VLM Self-Improvement via Strategic Gamified Self-Play
• From Perception to Cognition: A Survey of Vision-Language Interactive Reasoning in Multimodal Large Language Models
• OIG-Bench: A Multi-Agent Annotated Benchmark for Multimodal One-Image Guides Understanding
• ZOO-Prune: Training-Free Token Pruning via Zeroth-Order Gradient Estimation in Vision-Language Models
• AstroMMBench: A Benchmark for Evaluating Multimodal Large Language Models Capabilities in Astronomy
• When MLLMs Meet Compression Distortion: A Coding Paradigm Tailored to MLLMs
• FreeRet: MLLMs as Training-Free Retrievers
• Unlocking Zero-Shot Geospatial Reasoning via Indirect Rewards
• VisualOverload: Probing Visual Understanding of VLMs in Really Dense Scenes
• HiDe: Rethinking The Zoom-IN method in High Resolution MLLMs via Hierarchical Decoupling
• RAVEN: Resilient Aerial Navigation via Open-Set Semantic Memory and Behavior Adaptation
• Decoupling Reasoning and Perception: An LLM-LMM Framework for Faithful Visual Reasoning
• GUI-PRA: Process Reward Agent for GUI Tasks
• Understanding Language Prior of LVLMs by Contrasting Chain-of-Embedding
• XGC-AVis: Towards Audio-Visual Content Understanding with a Multi-Agent Collaborative System
• MMPB: It's Time for Multi-Modal Personalization
• Color Names in Vision-Language Models
• Explaining multimodal LLMs via intra-modal token interactions
• InfiMed-Foundation: Pioneering Advanced Multimodal Medical Models with Compute-Efficient Pre-Training and Multi-Stage Fine-Tuning
• UrbanFeel: A Comprehensive Benchmark for Temporal and Perceptual Understanding of City Scenes through Human Perspective
• WAVE: Learning Unified & Versatile Audio-Visual Embeddings with Multimodal LLM
• From Bias to Balance: Exploring and Mitigating Spatial Bias in LVLMs
• RISK: A Framework for GUI Agents in E-commerce Risk Management
• Mixture-of-Visual-Thoughts: Exploring Context-Adaptive Reasoning Mode Selection for General Visual Reasoning
• Visual Multi-Agent System: Mitigating Hallucination Snowballing via Visual Flow
• Training-Free Multimodal Deepfake Detection via Graph Reasoning
• MMR1: Enhancing Multimodal Reasoning with Variance-Aware Sampling and Open Resources
• ArchGPT: Understanding the World's Architectures with Large Multimodal Models
• CaTS-Bench: Can Language Models Describe Numeric Time Series?
• OmniScene: Attention-Augmented Multimodal 4D Scene Understanding for Autonomous Driving
• FreezeVLA: Action-Freezing Attacks against Vision-Language-Action Models
• Calibrate Before Reason: Robust Visual Token Reduction against Semantic Drift in VLMs
• One Patch Is Enough: Reinforcement-Optimized Visual Token Grounding for MLLM-Based Scene Text Spotting
• VisualRouter: Query-Grounded Visual Sampling for Long Video Understanding
• PanDent: Toward Comprehensive Tooth-Level Structure-Language Consistency in Dental Radiology
• QCalEval: Benchmarking Vision-Language Models for Quantum Calibration Plot Understanding
• S-GRPO: Unified Post-Training for Large Vision-Language Models
• Harm or Humor: A Multimodal, Multilingual Benchmark for Overt and Covert Harmful Humor
• ConViS-Bench: Estimating Video Similarity Through Semantic Concepts
• Citrus-V: Advancing Medical Foundation Models with Unified Medical Image Grounding for Clinical Reasoning
• How Far are VLMs from Visual Spatial Intelligence? A Benchmark-Driven Perspective
• MVT: Mask-Grounded Vision-Language Models for Taxonomy-Aligned Land-Cover Tagging
• UniECG: Understanding and Generating ECG in One Unified Model
• VIR-Bench: Evaluating Geospatial and Temporal Understanding of MLLMs via Travel Video Itinerary Reconstruction
• SD-VLM: Spatial Measuring and Understanding with Depth-Encoded Vision-Language Models
• TempSamp-R1: Effective Temporal Sampling with Reinforcement Fine-Tuning for Video LLMs
• The 1st Solution for 7th LSVOS RVOS Track: SaSaSa2VA
• Catching the Details: Self-Distilled RoI Predictors for Fine-Grained MLLM Perception
• Are VLMs Ready for Lane Topology Awareness in Autonomous Driving?
• MANZANO: A Simple and Scalable Unified Multimodal Model with a Hybrid Vision Tokenizer
• See&Trek: Training-Free Spatial Prompting for Multimodal Large Language Model
• Qianfan-VL: Domain-Enhanced Universal Vision-Language Models
• BTL-UI: Blink-Think-Link Reasoning Model for GUI Agent
• BaseReward: A Strong Baseline for Multimodal Reward Model
• ORIC: Benchmarking Object Recognition under Contextual Incongruity in Large Vision-Language Models
• SERVAL: Surprisingly Effective Zero-Shot Visual Document Retrieval Powered by Large Vision and Language Models
• RynnVLA-001: Using Human Demonstrations to Improve Robot Manipulation
• Embodied Arena: A Comprehensive, Unified, and Evolving Evaluation Platform for Embodied AI
• Frame Sampling Strategies Matter: A Benchmark for small vision language models
• Can Current AI Models Count What We Mean, Not What They See? A Benchmark and Systematic Evaluation
• AdaThinkDrive: Adaptive Thinking via Reinforcement Learning for Autonomous Driving
• SAIL-VL2 Technical Report
• Baseer: A Vision-Language Model for Arabic Document-to-Markdown OCR
• LLM-I: LLMs are Naturally Interleaved Multimodal Creators
• MARS2 2025 Challenge on Multimodal Reasoning: Datasets, Methods, Results, Discussion, and Outlook
• The LLM Already Knows: Estimating LLM-Perceived Question Difficulty via Hidden Representations
• The Art of Saying "Maybe": A Conformal Lens for Uncertainty Benchmarking in VLMs
• MiniCPM-V 4.5: Cooking Efficient MLLMs via Architecture, Data, and Training Recipe
• Igniting VLMs toward the Embodied Space
• MindVL: Towards Efficient and Effective Training of Multimodal Large Language Models on Ascend NPUs
• Look Again, Think Slowly: Enhancing Visual Reflection in Vision-Language Models
• Adapting and Evaluating Multimodal Large Language Models for Adolescent Idiopathic Scoliosis Self-Management: A Divide and Conquer Framework
• Enhancing Generalization in Vision-Language-Action Models by Preserving Pretrained Representations
• MagicMirror: A Large-Scale Dataset and Benchmark for Fine-Grained Artifacts Assessment in Text-to-Image Generation
• InPhyRe Discovers: Large Multimodal Models Struggle in Inductive Physical Reasoning
• Measuring Epistemic Humility in Multimodal Large Language Models
• TextlessRAG: End-to-End Visual Document RAG by Speech Without Text
• Toward Socially Aware Vision-Language Models: Evaluating Cultural Competence Through Multimodal Story Generation
• GLEAM: Learning to Match and Explain in Cross-View Geo-Localization
• Video Parallel Scaling: Aggregating Diverse Frame Subsets for VideoLLMs
• Reconstruction Alignment Improves Unified Multimodal Models
• Text4Seg++: Advancing Image Segmentation via Generative Language Modeling
• PictOBI-20k: Unveiling Large Multimodal Models in Visual Decipherment for Pictographic Oracle Bone Characters
• DreamPRM-1.5: Unlocking the Potential of Each Instance for Multimodal Process Reward Model Training
• FPC-VLA: A Vision-Language-Action Framework with a Supervisor for Failure Prediction and Correction
• Skywork UniPic 2.0: Building Kontext Model with Online RL for Unified Multimodal Model
• E-ARMOR: Edge case Assessment and Review of Multilingual Optical Character Recognition
• OneCAT: Decoder-Only Auto-Regressive Model for Unified Understanding and Generation
• ProMQA-Assembly: Multimodal Procedural QA Dataset on Assembly
• Easier Painting Than Thinking: Can Text-to-Image Models Set the Stage, but Not Direct the Play?
• VLMs-in-the-Wild: Bridging the Gap Between Academic Benchmarks and Enterprise Reality
• Omnidirectional Spatial Modeling from Correlated Panoramas
• RSCC: A Large-Scale Remote Sensing Change Caption Dataset for Disaster Events
• PixFoundation 2.0: Do Video Multi-Modal LLMs Use Motion in Visual Grounding?
• Kwai Keye-VL 1.5 Technical Report
• Measuring Image-Relation Alignment: Reference-Free Evaluation of VLMs and Synthetic Pre-training for Open-Vocabulary Scene Graph Generation
• Less Redundancy: Boosting Practicality of Vision Language Model in Walking Assistants
• Variation-aware Vision Token Dropping for Faster Large Vision-Language Models
• CogDriver: Integrating Cognitive Inertia for Temporally Coherent Planning in Autonomous Driving
• VideoRewardBench: Comprehensive Evaluation of Multimodal Reward Models for Video Understanding
• Safe-LLaVA: A Privacy-Preserving Vision-Language Dataset and Benchmark for Biometric Safety
• Is this chart lying to me? Automating the detection of misleading visualizations
• ELV-Halluc: Benchmarking Semantic Aggregation Hallucinations in Long Video Understanding
• R-4B: Incentivizing General-Purpose Auto-Thinking Capability in MLLMs via Bi-Mode Annealing and Reinforce Learning
• Intern-S1: A Scientific Multimodal Foundation Model
• Veritas: Generalizable Deepfake Detection via Pattern-Aware Reasoning
• SWIRL: A Staged Workflow for Interleaved Reinforcement Learning in Mobile GUI Control
• GLSim: Detecting Object Hallucinations in LVLMs via Global-Local Similarity
• Video-LevelGauge: Investigating Contextual Positional Bias in Large Video Language Models
• Mind the Third Eye! Benchmarking Privacy Awareness in MLLM-powered Smartphone Agents
• CrossHOI-Bench: A Unified Benchmark for HOI Evaluation across Vision-Language Models and HOI-Specific Methods
• PRISM: Robust VLM Alignment with Principled Reasoning for Integrated Safety in Multimodality
• InternVL3.5: Advancing Open-Source Multimodal Models in Versatility, Reasoning, and Efficiency
• SEAM: Semantically Equivalent Across Modalities Benchmark for Vision-Language Models
• SpotEdit: Evaluating Visually-Guided Image Editing Methods
• MedRepBench: A Comprehensive Benchmark for Medical Report Interpretation
• Mobile-Agent-v3: Fundamental Agents for GUI Automation
• MME-SCI: A Comprehensive and Challenging Science Benchmark for Multimodal Large Language Models
• HumanPCR: Probing MLLM Capabilities in Diverse Human-Centric Scenes
• AdaDocVQA: Adaptive Framework for Long Document Visual Question Answering in Low-Resource Settings
• STER-VLM: Spatio-Temporal With Enhanced Reference Vision-Language Models
• DianJin-OCR-R1: Enhancing OCR Capabilities via a Reasoning-and-Tool Interleaved Vision-Language Model
• Vision-G1: Towards General Vision Language Reasoning with Multi-Domain Data Curation
• Adaptive Reinforcement for Open-ended Medical Reasoning via Semantic-Guided Reward Collapse Mitigation
• ViDA-UGC: Detailed Image Quality Analysis via Visual Distortion Assessment for UGC Images
• MOON: Generative MLLM-based Multimodal Representation Learning for E-commerce Product Understanding
• Thyme: Think Beyond Images
• Ovis2.5 Technical Report
• EgoCross: Benchmarking Multimodal Large Language Models for Cross-Domain Egocentric Video Question Answering
• HumanSense: From Multimodal Perception to Empathetic Context-Aware Responses through Reasoning MLLMs
• A Study of Commonsense Reasoning over Visual Object Properties
• We-Math 2.0: A Versatile MathBook System for Incentivizing Visual Mathematical Reasoning
• Empowering Multimodal LLMs with External Tools: A Comprehensive Survey
• VisCodex: Unified Multimodal Code Generation via Merging Vision and Coding Models
• SHALE: A Scalable Benchmark for Fine-grained Hallucination Evaluation in LVLMs
• CARES: Collaborative Agentic Reasoning for Error Detection in Surgery
• Scaling Learned Image Compression Models up to 1 Billion
• Bridging Formal Language with Chain-of-Thought Reasoning to Geometry Problem Solving
• Towards Effective MLLM Jailbreaking Through Balanced On-Topicness and OOD-Intensity
• MedReasoner: Reinforcement Learning Drives Reasoning Grounding from Clinical Thought to Pixel-Level Precision
• RSVLM-QA: A Benchmark Dataset for Remote Sensing Vision Language Model-based Question Answering
• Find Them All: Unveiling MLLMs for Versatile Person Re-identification
• InfoCausalQA:Can Models Perform Non-explicit Causal Reasoning Based on Infographic?
• Affordance-R1: Reinforcement Learning for Generalizable Affordance Reasoning in Multimodal Large Language Model
• Bifrost-1: Bridging Multimodal LLMs and Diffusion Models with Patch-level CLIP Latents
• MathReal: We Keep It Real! A Real Scene Benchmark for Evaluating Math Reasoning in Multimodal Large Language Models
• VQAThinker: Exploring Generalizable and Explainable Video Quality Assessment via Reinforcement Learning
• Can Large Models Fool the Eye? A New Turing Test for Biological Animation
• Test-Time Reinforcement Learning for GUI Grounding via Region Consistency
• Shuffle-R1: Efficient RL framework for Multimodal Large Language Models via Data-centric Dynamic Shuffle
• Follow-Your-Instruction: A Comprehensive MLLM Agent for World Data Synthesis
• mKG-RAG: Multimodal Knowledge Graph-Enhanced RAG for Visual Question Answering
• InfiGUI-G1: Advancing GUI Grounding with Adaptive Exploration Policy Optimization
• VER-Bench: Evaluating MLLMs on Reasoning with Fine-Grained Visual Evidence
• ConfProBench: A Confidence Evaluation Benchmark for MLLM-Based Process Judges
• Knowledge to Sight: Reasoning over Visual Attributes via Knowledge Decomposition for Abnormality Grounding
• PET2Rep: Towards Vision-Language Model-Drived Automated Radiology Report Generation for Positron Emission Tomography
• Static and Plugged: Make Embodied Evaluation Simple
• Beyond the Visible: Benchmarking Occlusion Perception in Multimodal Large Language Models
• SAM2-UNeXT: An Improved High-Resolution Baseline for Adapting Foundation Models to Downstream Segmentation Tasks
• VLMQ: Efficient Post-Training Quantization for Large Vision-Language Models via Hessian Augmentation
• Geoint-R1: Formalizing Multimodal Geometric Reasoning with Dynamic Auxiliary Constructions
• AVATAR: Reinforcement Learning to See, Hear, and Reason Over Video
• Refine-IQA: Multi-Stage Reinforcement Finetuning for Perceptual Image Quality Assessment
• Engagement Prediction of Short Videos with Large Multimodal Models
• Mapillary Vistas Validation for Fine-Grained Traffic Signs: A Benchmark Revealing Vision-Language Model Limitations
• A Glimpse to Compress: Dynamic Visual Token Pruning for Large Vision-Language Models
• MagicVL-2B: Empowering Vision-Language Models on Mobile Devices with Lightweight Visual Encoders via Curriculum Learning
• What Makes "Good" Distractors for Object Hallucination Evaluation in Large Vision-Language Models?
• MAP: Mitigating Hallucinations in Large Vision-Language Models with Map-Level Attention Processing
• RoboMemory: A Brain-inspired Multi-memory Agentic Framework for Interactive Environmental Learning in Physical Embodied Systems
• SpectrumWorld: Artificial Intelligence Foundation for Spectroscopy
• CausalStep: A Benchmark for Explicit Stepwise Causal Reasoning in Videos
• Bidirectional Likelihood Estimation with Multi-Modal Large Language Models for Text-Video Retrieval
• UniLiP: Adapting CLIP for Unified Multimodal Understanding, Generation and Editing
• Being-H0: Vision-Language-Action Pretraining from Large-Scale Human Videos
• VL-Cogito: Progressive Curriculum Reinforcement Learning for Advanced Multimodal Reasoning
• FairReason: Balancing Reasoning and Social Bias in MLLMs
• Adapting Vehicle Detectors for Aerial Imagery to Unseen Domains with Weak Supervision
• Learning Only with Images: Visual Reinforcement Learning with Reasoning, Rendering, and Visual Feedback
• A Survey of Token Compression for Efficient Multimodal Large Language Models
• Predicting Brain Responses To Natural Movies With Multimodal LLMs
• Position: Reasoning After Perception Means Reasoning Without Vision
• MMBench-GUI: Hierarchical Multi-Platform Evaluation Framework for GUI Agents
• OS-MAP: How Far Can Computer-Using Agents Go in Breadth and Depth?
• IntentVCNet: Bridging Spatio-Temporal Gaps for Intention-Oriented Controllable Video Captioning
• Chart-R1: Chain-of-Thought Supervision and Reinforcement for Advanced Chart Reasoner
• EgoExoBench: A Benchmark for First- and Third-person View Video Understanding in MLLMs
• Seeing is Free, Speaking is Not: Uncovering the True Energy Bottleneck in Edge VLM Inference
• U-MARVEL: Unveiling Key Factors for Universal Multimodal Retrieval via Embedding Learning with MLLMs
• PIXELRAG: Web Screenshots Beat Text for Retrieval-Augmented Generation
• BusterX++: Towards Unified Cross-Modal AI-Generated Content Detection and Explanation with MLLM
• ArtiMuse: Fine-Grained Image Aesthetics Assessment with Joint Scoring and Expert-Level Understanding
• ThinkAct: Vision-Language-Action Reasoning via Reinforced Visual Latent Planning
• Edge-case Synthesis for Fisheye Object Detection: A Data-centric Perspective
• AD²-Bench: A Hierarchical CoT Benchmark for MLLM in Autonomous Driving under Adverse Conditions
• Describe Anything Model for Visual Question Answering on Text-rich Images
• MindJourney: Test-Time Scaling with World Models for Spatial Reasoning
• UGC-VideoCaptioner: An Omni UGC Video Detail Caption Model and New Benchmarks
• FaceLLM: A Multimodal Large Language Model for Face Understanding
• EmbRACE-3K: Embodied Reasoning and Action in Complex Environments
• VRU-Accident: A Vision-Language Benchmark for Video Question Answering and Dense Captioning for Accident Scene Understanding
• RAMA: Retrieval-Augmented Multi-Agent Framework for Misinformation Detection in Multimodal Fact-Checking
• Multilingual Multimodal Software Developer for Code Generation
• M2-Reasoning: Empowering MLLMs with Unified General and Spatial Reasoning
• Scaling RL to Long Videos
• Traceable Evidence Enhanced Visual Grounded Reasoning: Evaluation and Methodology
• Video-RTS: Rethinking Reinforcement Learning and Test-Time Scaling for Efficient and Enhanced Video Reasoning
• Perception-Aware Policy Optimization for Multimodal Reasoning
• Skywork-R1V3 Technical Report
• BlueLM-2.5-3B Technical Report
• Open Vision Reasoner: Transferring Linguistic Cognitive Behavior for Visual Reasoning
• VectorLLM: Human-like Extraction of Structured Building Contours vis Multimodal LLMs
• ZERO: Industry-ready Vision Foundation Model with Multi-modal Prompts
• Anti-Prompt: Image Protection against Text-Guided Image-to-Video Generation
• Foundation versus Domain-specific Models: Performance Comparison, Fusion, and Explainability in Face Recognition
• ConceptMix++: Leveling the Playing Field in Text-to-Image Benchmarking via Iterative Prompt Optimization
• BMMR: A Large-Scale Bilingual Multimodal Multi-Discipline Reasoning Dataset
• VLMs have Tunnel Vision: Evaluating Nonlocal Visual Reasoning in Leading VLMs
• TokenSwap: Benchmarking and Reducing the Modality Gap in Multimodal LLMs
• Kwai Keye-VL Technical Report
• SAILViT: Towards Robust and Generalizable Visual Backbones for MLLMs via Gradual Feature Refinement
• GLM-4.5V and GLM-4.1V-Thinking: Towards Versatile Multimodal Reasoning with Scalable Reinforcement Learning
• Improving the Reasoning of Multi-Image Grounding in MLLMs via Reinforcement Learning
• M3MAD-Bench: Multi-Dimensional Evaluation of Multi-Agent Debate Across Domains and Modalities
• Just Noticeable Difference for Large Multimodal Models
• DenseWorld-1M: Towards Detailed Dense Grounded Caption in the Real World
• ZonUI-3B: A Lightweight Vision-Language Model for Cross-Resolution GUI Grounding
• CyberV: Cybernetics for Test-time Scaling in Video Understanding
• IR3D-Bench: Evaluating Vision-Language Model Scene Understanding as Agentic Inverse Rendering
• Ovis-U1 Technical Report
• Towards Explainable Bilingual Multimodal Misinformation Detection and Localization
• MARBLE: A Hard Benchmark for Multimodal Spatial Reasoning and Planning
• HAIBU-ReMUD: Reasoning Multimodal Ultrasound Dataset and Model Bridging to General Specific Domains
• Test-Time Consistency in Vision Language Models
• Can Video Large Multimodal Models Think Like Doubters-or Double-Down: A Study on Defeasible Video Entailment
• Rethinking Visual Token Reduction in LVLMs Under Cross-Modal Misalignment
• Remote Sensing Large Vision-Language Model: Semantic-augmented Multi-level Alignment and Semantic-aware Expert Modeling
• R1-Track: Direct Application of MLLMs to Visual Object Tracking via Reinforcement Learning
• ImplicitQA: Going beyond frames towards Implicit Video Reasoning
• Task-Aware KV Compression For Cost-Effective Long Video Understanding
• BitVLA: 1-bit Vision-Language-Action Models for Robotics Manipulation
• DeepVideo-R1: Video Reinforcement Fine-Tuning via Difficulty-aware Regressive GRPO
• LLaVA-Pose: Enhancing Human Pose and Action Understanding via Keypoint-Integrated Instruction Tuning
• World-aware Planning Narratives Enhance Large Vision-Language Model Planner
• ShotBench: Expert-Level Cinematic Understanding in Vision-Language Models
• Seeing is Believing? Mitigating OCR Hallucinations in Multimodal Large Language Models
• Vision Transformers Don't Need Trained Registers
• Fine-grained Token Allocation Via Operation Pruning for Efficient MLLMs
• GUI-Reflection: Empowering Multimodal GUI Models with Self-Reflection Behavior
• MedTVT-R1: A Multimodal LLM Empowering Medical Reasoning and Diagnosis
• CLGRPO: Reasoning Ability Enhancement for Small VLMs
• PostAlign: Multimodal Grounding as a Corrective Lens for MLLMs
• PhysUniBench: A Multi-Modal Physics Reasoning Benchmark at Undergraduate Level
• CLiViS: Unleashing Cognitive Map through Linguistic-Visual Synergy for Embodied Visual Reasoning
• UniFork: Exploring Modality Alignment for Unified Multimodal Understanding and Generation
• GRPO-CARE: Consistency-Aware Reinforcement Learning for Multimodal Reasoning
• Can Generalist Vision Language Models (VLMs) Rival Specialist Medical VLMs? Benchmarking and Strategic Insights
• SpaCE-10: A Comprehensive Benchmark for Multimodal Large Language Models in Compositional Spatial Intelligence
• ReCogDrive: A Reinforced Cognitive Framework for End-to-End Autonomous Driving
• Recognition through Reasoning: Reinforcing Image Geo-localization with Large Vision-Language Models
• StreetLens: Enabling Human-Centered AI Agents for Neighborhood Assessment from Street View Imagery
• HKD4VLM: A Progressive Hybrid Knowledge Distillation Framework for Robust Multimodal Hallucination and Factuality Detection in VLMs
• AceReason-Nemotron 1.1: Advancing Math and Code Reasoning through SFT and RL Synergy
• Metis-RISE: RL Incentivizes and SFT Enhances Multimodal Reasoning Model Learning
• Native Visual Understanding: Resolving Resolution Dilemmas in Vision-Language Models
• Guiding Cross-Modal Representations with MLLM Priors via Preference Alignment
• Image Corruption-Inspired Membership Inference Attacks against Large Vision-Language Models
• Unsupervised Document and Template Clustering using Multimodal Embeddings
• Dynamic Mixture of Curriculum LoRA Experts for Continual Multimodal Instruction Tuning
• Aligning MLLM Benchmark With Human Preferences via Structural Equation Modeling
• VGR: Visual Grounded Reasoning
• Benchmarking Multimodal LLMs on Recognition and Understanding over Chemical Tables
• VFaith: Do Large Multimodal Models Really Reason on Seen Images Rather than Previous Memories?
• Breaking Bad Molecules: Are MLLMs Ready for Structure-Level Molecular Detoxification?
• Beyond Attention or Similarity: Maximizing Conditional Diversity for Token Pruning in MLLMs
• Scientists' First Exam: Probing Cognitive Abilities of MLLM via Perception, Understanding, and Reasoning
• PuzzleWorld: A Benchmark for Multimodal, Open-Ended Reasoning in Puzzlehunts
• VideoChat-A1: Thinking with Long Videos by Chain-of-Shot Reasoning
• Movie Facts and Fibs (MF²): A Benchmark for Long Movie Understanding
• VideoMolmo: Spatio-Temporal Grounding Meets Pointing
• Unleashing Hour-Scale Video Training for Long Video-Language Understanding
• AV-Reasoner: Improving and Benchmarking Clue-Grounded Audio-Visual Counting for MLLMs
• SIV-Bench: A Video Benchmark for Social Interaction Understanding and Reasoning
• Reasoning-Aligned Perception Decoupling for Scalable Multi-modal Reasoning
• VideoMathQA: Benchmarking Mathematical Reasoning via Multimodal Understanding in Videos
• Degradation-Aware Image Enhancement via Vision-Language Classification
• From Objects to Anywhere: A Holistic Benchmark for Multi-level Visual Grounding in 3D Scenes
• LUT: Latent Utility Training for Visual Reasoning
• GeoArbiter: Verifiability-Guided Grounding for Remote-Sensing Multimodal LLMs
• VLMs Can Aggregate Scattered Training Patches
• Rex-Thinker: Grounded Object Referring via Chain-of-Thought Reasoning
• Evaluating MLLMs with Multimodal Multi-image Reasoning Benchmark
• OmniSpatial: Towards Comprehensive Spatial Reasoning Benchmark for Vision Language Models
• VS-Bench: Evaluating VLMs for Strategic Abilities in Multi-Agent Environments
• SVGenius: Benchmarking LLMs in SVG Understanding, Editing and Generation
• SurgVLM: A Large Vision-Language Model and Systematic Evaluation Benchmark for Surgical Intelligence
• MotionSight: Boosting Fine-Grained Motion Understanding in Multimodal LLMs
• Zoom-Refine: Boosting High-Resolution Multimodal Understanding via Localized Zoom and Self-Refinement
• FLEX: A Largescale Multimodal, Multiview Dataset for Learning Structured Representations for Fitness Action Quality Assessment
• ReFoCUS: Reinforcement-guided Frame Optimization for Contextual Understanding
• Affordance Benchmark for MLLMs
• GThinker: Towards General Multimodal Reasoning via Cue-Guided Rethinking
• Infinity Parser: Layout Aware Reinforcement Learning for Scanned Document Parsing
• MedBookVQA: A Systematic and Comprehensive Medical Benchmark Derived from Open-Access Book
• Chain-of-Frames: Advancing Video Understanding in Multimodal LLMs via Frame-Aware Reasoning
• The LLM Bottleneck: Why Open-Source Vision LLMs Struggle with Hierarchical Visual Recognition
• Draw ALL Your Imagine: A Holistic Benchmark and Agent Framework for Complex Instruction-based Image Generation
• Benchmarking Foundation Models for Zero-Shot Biometric Tasks
• CSVQA: A Chinese Multimodal Benchmark for Evaluating STEM Reasoning Capabilities of VLMs
• Mixed-R1: Unified Reward Perspective For Reasoning Capability in Multimodal Large Language Models
• Seeing is Not Reasoning: MVPBench for Graph-based Evaluation of Multi-path Visual Physical CoT
• ScaleLong: A Multi-Timescale Benchmark for Long Video Understanding
• OmniEarth-Bench: Towards Holistic Evaluation of Earth's Six Spheres and Cross-Spheres Interactions with Multimodal Observational Earth Data
• VideoReasonBench: Can MLLMs Perform Vision-Centric Complex Video Reasoning?
• InfiMed: Low-Resource Medical MLLMs with Advancing Understanding and Reasoning
• Generalizable Video Quality Assessment via Weak-to-Strong Learning
• SNS-Bench-VL: Benchmarking Multimodal Large Language Models in Social Networking Services
• VF-Eval: Evaluating Multimodal LLMs for Generating Feedback on AIGC Videos
• SPR-128K: A New Benchmark for Spatial Plausibility Reasoning with Multimodal Large Language Models
• MMSI-Bench: A Benchmark for Multi-Image Spatial Intelligence
• mRAG: Elucidating the Design Space of Multi-modal Retrieval-Augmented Generation
• DIP-R1: Deep Inspection and Perception with RL Looking Through and Understanding Complex Scenes
• OpenUni: A Simple Baseline for Unified Multimodal Understanding and Generation
• Training Free Stylized Abstraction
• Towards Comprehensive Scene Understanding: Integrating First and Third-Person Views for LVLMs
• Sherlock: Self-Correcting Reasoning in Vision-Language Models
• Improve Multi-Modal Embedding Learning via Explicit Hard Negative Gradient Amplifying
• MSEarth: A Multimodal Scientific Dataset and Benchmark for Phenomena Uncovering in Earth Science
• ViewSpatial-Bench: Evaluating Multi-perspective Spatial Localization in Vision-Language Models
• MME-VideoOCR: Evaluating OCR-Based Capabilities of Multimodal LLMs in Video Scenarios
• Video-Holmes: Can MLLM Think Like Holmes for Complex Video Reasoning?
• DisasterM3: A Remote Sensing Vision-Language Dataset for Disaster Damage Assessment and Response
• DriveRX: A Vision-Language Reasoning Model for Cross-Task Autonomous Driving
• DynamicVL: Benchmarking Multimodal Large Language Models for Dynamic City Understanding
• UI-Genie: A Self-Improving Approach for Iteratively Boosting MLLM-based Mobile GUI Agents
• GeoLLaVA-8K: Scaling Remote-Sensing Multimodal Large Language Models to 8K Resolution
• Vad-R1: Towards Video Anomaly Reasoning via Perception-to-Cognition Chain-of-Thought
• Seeing is Believing, but How Much? A Comprehensive Analysis of Verbalized Calibration in Vision-Language Models
• MMPerspective: Do MLLMs Understand Perspective? A Comprehensive Benchmark for Perspective Perception, Reasoning, and Robustness
• Modality Curation: Building Universal Embeddings for Advanced Multimodal Information Retrieval
• R1-Reward: Training Multimodal Reward Model Through Stable Reinforcement Learning
• Diagnosing and Mitigating Modality Interference in Multimodal Large Language Models
• Benchmarking Multimodal Knowledge Conflict for Large Multimodal Models
• Shifting AI Efficiency From Model-Centric to Data-Centric Compression
• SATORI-R1: Incentivizing Multimodal Reasoning through Explicit Visual Anchoring
• InfoChartQA: A Benchmark for Multimodal Question Answering on Infographic Charts
• ToDRE: Effective Visual Token Pruning via Token Diversity and Task Relevance
• GRE Suite: Geo-localization Inference via Fine-Tuned Vision-Language Models and Enhanced Reasoning Chains
• So-Fake: Benchmarking and Explaining Social Media Image Forgery Detection
• MLLMs are Deeply Affected by Modality Bias
• Generative RLHF-V: Learning Principles from Multi-modal Human Preference
• ReasonMap: Towards Fine-Grained Visual Reasoning from Transit Maps
• DeepVoyager-VL: Incentivizing Vision-in-the-Loop Search for Long-Horizon Multimodal Agents
• Illuminating Visual Identity in Universal Multimodal Embeddings
• Co-Reinforcement Learning for Unified Multimodal Understanding and Generation
• VIBE: Annotation-Free Video-to-Text Information Bottleneck Evaluation for TL;DR
• FullFront: Benchmarking MLLMs Across the Full Front-End Engineering Workflow
• VC-Tooler: Learning Compositional and Adaptive Visual Tool Use
• Let Androids Dream of Electric Sheep: A Human-Inspired Image Implication Understanding and Reasoning Framework
• DetailMaster: Can Your Text-to-Image Model Handle Long Prompts?
• OCR-Reasoning Benchmark: Unveiling the True Capabilities of MLLMs in Complex Text-Rich Image Reasoning
• RBench-V: A Primary Assessment for Visual Reasoning Models with Multi-modal Outputs
• Bridging the Dynamic Perception Gap: Training-Free Draft Chain-of-Thought for Dynamic Multimodal Spatial Reasoning
• Benchmarking Retrieval-Augmented Multimodal Generation for Document Question Answering
• IRONIC: Coherence-Aware Reasoning Chains for Multi-Modal Sarcasm Detection
• QuickVideo: Real-Time Long Video Understanding with System Algorithm Co-Design
• SpatialScore: Towards Comprehensive Evaluation for Spatial Intelligence
• STAR-R1: Spatial TrAnsformation Reasoning by Reinforcing Multimodal LLMs
• Better Safe Than Sorry? Overreaction Problem of Vision Language Models in Visual Emergency Recognition
• InstructSAM: A Training-Free Framework for Instruction-Oriented Remote Sensing Object Recognition
• LENS: Multi-level Evaluation of Multimodal Reasoning with Large Language Models
• CineTechBench: A Benchmark for Cinematographic Technique Understanding and Generation
• GRIT: Teaching MLLMs to Think with Images
• MIEScore: Human-Aligned Evaluation for Multi-Source Image Editing
• FAU at ImageCLEF 2026 Task on Multimodal Reasoning Robust Candidate Scoring and Concise Multilingual Visual Answering
• Distill What the Student Can See: Fisher-Projected On-Policy Distillation for Vision-Language Models
• VaRS-Doc: Interpretation-Aware Variant Representations via Latent Self-Probing for Visual Document Retrieval
• VoQA: Visual-only Question Answering
• UniGen: Enhanced Training & Test-Time Strategies for Unified Multimodal Understanding and Generation
• KORGym: A Dynamic Game Platform for LLM Reasoning Evaluation
• MonitorVLM-v2: A Deployed Vision-Language Framework for Real-Time Safety Violation Detection
• DeepEyes: Incentivizing "Thinking with Images" via Reinforcement Learning
• ViC-Bench: Benchmarking Visual-Interleaved Chain-of-Thought Capability in MLLMs with Free-Style Intermediate State Representations
• VideoEval-Pro: Robust and Realistic Long Video Understanding Evaluation
• PlanGPT-VL: Enhancing Urban Planning with Domain-Specific Vision-Language Models
• UniHEAR: Unified Heterogeneous-Source Attentive Retrieval for Knowledge-Based Visual Question Answering
• ViPlan: A Benchmark for Visual Planning with Symbolic Predicates and Vision-Language Models
• LogicOCR: Do Your Large Multimodal Models Excel at Logical Reasoning on Text-Rich Images?
• CorBenchX: Large-Scale Chest X-Ray Error Dataset and Vision-Language Model Benchmark for Report Error Correction
• MedSG-Bench: A Benchmark for Medical Image Sequences Grounding
• Location-Aware Fine-Grained Representation Learning for Medical Vision Foundation Models
• Human-Aligned Bench: Fine-Grained Assessment of Reasoning Ability in MLLMs vs. Humans
• Decoding Children's Gait Behavior
• VISTA: Enhancing Vision-Text Alignment in MLLMs via Cross-Modal Mutual Information Maximization
• MIRAGE: A Multi-modal Benchmark for Spatial Perception, Reasoning, and Intelligence
• SafeBuild-Bench: A Temporal-Robust Construction Safety Benchmark with Graph-Enhanced Data Mining
• Linguistic Context Recodes Visual Representations in Vision-Language Models
• CoLT: Teaching Multi-Modal Models to Think with Chain of Latent Thoughts
• EgoIntent: A Pre-Outcome Micro-Step Benchmark for Understanding What, Why, and Next
• DriveSOTIF: Advancing Perception SOTIF Through Multimodal Large Language Models
• EmoScene: A Dual-space Dataset for Controllable Affective Image Generation
• NarrativeTrack: Evaluating Entity-Centric Reasoning for Narrative Understanding
• Steerable Visual Representations
• UniDriveVLA: Unifying Understanding, Perception, and Action Planning for Autonomous Driving
• VideoHallu: Evaluating and Mitigating Multi-modal Hallucinations on Synthetic Video Understanding
• Kwai Keye-VL-2.0 Technical Report
• SpaceMind++: Toward Allocentric Cognitive Maps for Spatially Grounded Video MLLMs
• Better, Stronger, Faster, and Broader: Structured All-Mask Prediction for MLLM-Based Segmentation
• FireRed-OCR Technical Report
• Children's Intelligence Tests Pose Challenges for MLLMs? KidGym: A 2D Grid-Based Reasoning Benchmark for MLLMs
• Beyond Language Modeling: An Exploration of Multimodal Pretraining
• Mobile-Agent-v3.5: Multi-platform Fundamental GUI Agents
• GDI-Bench: A Benchmark for General Document Intelligence with Vision and Reasoning Decoupling
• Reinforced MLLM: A Survey on RL-Based Reasoning in Multimodal Large Language Models
• Robust Grounding with MLLMs Against Occlusion and Small Objects via Language-Guided Semantic Cues
• Act Wisely: Cultivating Meta-Cognitive Tool Use in Agentic Multimodal Models
• Reinforced Attention Learning
• Caved or Convinced: Temporal Sampling Gates Claim Deference in Video Large Language Models
• Adaptive Two-Stage Visual Token Pruning for Efficient Inference in Video-Language Models
• DocTrace: Towards Traceable Long Document VQA via Hierarchical Evidence Graph Reasoning
• A Review of 3D Object Detection with Vision-Language Models
• REZE: Recognition-Based Zero-Shot Extraction for Video Temporal Grounding
• Representing Visual Evidence for Item Difficulty Prediction: Visual Textualization and Image-Native Modeling
• DIVE: Dynamic Iterative Visual Evidence Construction for Efficient Vision-Language Models
• When Prompts Become Pixels: Prompt-Region Grounding for Multimodal Reasoning
• Teaching Foundation Models to Read mmWave: Pose-Guided Kinematic Representation for Human Behavior Understanding
• MOON3.0: Reasoning-aware Multimodal Representation Learning for E-commerce Product Understanding
• Reasoning Dynamics and the Limits of Monitoring Modality Reliance in Vision-Language Models
• Skywork R1V2: Multimodal Hybrid Reinforcement Learning for Reasoning
• TAU-Bench: From Anomaly Instance Tracking to Fine-Grained Video Anomaly Understanding
• HunyuanOCR-1.5: Making Lightweight OCR VLMs Faster and Better
• Mavors: Multi-granularity Video Representation for Multimodal Large Language Model
• SFT or RL? An Early Investigation into Training R1-Like Reasoning Large Vision-Language Models
• V-MAGE: A Game Evaluation Framework for Assessing Vision-Centric Capabilities in Multimodal Large Language Models

Related

• Mono-InternVL: Pushing the Boundaries of Monolithic Multimodal Large Language Models with Endogenous Visual Pre-training
• Expanding Performance Boundaries of Open-Source Multimodal Models with Model, Data, and Test-Time Scaling
• Mono-InternVL-1.5: Towards Cheaper and Faster Monolithic Multimodal Large Language Models
• How Far Are We to GPT-4V? Closing the Gap to Commercial Multimodal Models with Open-Source Suites
• MindGPT-4ov: An Enhanced MLLM via a Multi-Stage Post-Training Paradigm