VLM Reasoning

VLM: Vision-Language Model

Overview

This field studies how models connect images or video with language to solve tasks involving spatial relations, time, causality, diagrams, and action. Common approaches include chain-of-thought prompting, visual tools and retrieval, structured representations such as scene graphs, and fine-tuning with process-level rewards. Benchmarks increasingly test whether answers and intermediate steps are grounded in visual evidence, including in multi-image, video, scientific, and embodied settings. A central tension is that longer reasoning can help on complex tasks but can also weaken visual grounding, amplify language-based shortcuts, or produce explanations that do not reflect how an answer was reached.

Latest papers

CardsList

875 papers20 papers in the last four weeks, down 82% on the four weeks before. 0.2% of all new papers.

  1. When Safety Rationales Hallucinate: Diagnosing and Grounding Multimodal Safety Guards

    Nov 19, 2026Yingshuo Wang, Weixiang Zhao, Guancheng Li +3MLLM SafetyVLM Hallucination Mitigation

  2. Teaching VLMs What to Say, Not How to Reason: Rethinking Counterfactual Reasoning in Autonomous Driving

    Nov 18, 2026Hayeon Oh, Donghwan LeeVLMs for Autonomous DrivingVLM Reasoning

  3. CanopyKV: Remember beyond the Crop for Fast Active Visual Reasoning

    Nov 18, 2026Lei Yang, Dayan Wu, Yiliang Zhu +6KV-Cache CompressionEfficient VLM Inference

  4. AVTR: Adaptive Multimodal Reasoning with Visual Imagination

    Nov 18, 2026Xiaoqian Shen, Yunyang Xiong, Changsheng Zhao +15Reasoning for VLMsVLM Reasoning

  5. BOS Summary: The First Token Matters

    Nov 17, 2026Hui Zhu, Saleh Zare Zade, Rafi Ibn Sultan +3Attention Intervention in VLMsAttention Sinks

  6. Visual Marking as a Native Reasoning Capability for Multimodal LLMs

    Nov 15, 2026Jiazhen Liu, Yanghao Wang, Chaolei Tan +1VLM ReasoningMultimodal Spatial Reasoning

  7. Negation Through Affirmation: Leveraging Positive Evidence in Vision-Language Models

    Nov 11, 2026Aashish Pokhrel, Tejas Anvekar, Shivanand Venkanna SheshappanavarVLM ReasoningVision-Language Retrieval

  8. AITHER: Authority-Informed Evidence Routing for Mixed-Authority Vision-Language Reasoning

    Nov 10, 2026Xinjin Li, Haoyue Zhang, Xinyi Gou +7VLM ReasoningVision-Language Grounding

  9. Viscot-steer: Disentangling and Transferring Visual Chain-of-Thought via Activation Steering

    Nov 8, 2026Ningning Wang, Beier Zhu, Xingyu Zhu +2Activation Steering in VLMsVLM Reasoning

  10. Multimodal Models Can See, But Can They Think? A Study of Turkish Visual Reasoning

    Nov 7, 2026Berke Bunyamin Sule, Yavuz Selim Aygan, Ahmed Zeer +6Reasoning for VLMsVLM Reasoning

  11. Reason First, Retrieve What Matters: Hypothesis-Guided Keyframe Selection for Long Video Understanding

    Nov 7, 2026Yisong Hu, Junbo Wang, Yuke Li +1Long-Video UnderstandingQuery-Aware Video Frame Selection

  12. Learning When to Think: Dual-Reference Offline Optimization for Adaptive VLM Reasoning

    Nov 6, 2026Hongbin Lin, Sizhe Zou, Juangui Xu +6Reasoning for VLMsVLM Reasoning

  13. VizDAGuide: Misleading Chart Question Answering in Compact Vision Language Models via DAG-Structured Supervision

    Nov 2, 2026Haiyang Zhang, Sicheng Song, Zixin Chen +4Chart QAVLM Reasoning

  14. Fewer Tokens, More Answers: Verifiable-Reward Tuning of an Ornith-Style VLM for Multi-View Driving QA

    Oct 31, 2026T N Rickesh, Pranav Vinodh, Sowmya Kamath SVLM ReasoningVisual Question Answering

  15. GRIT: Gradient Masking for Visually Misgrounded Tokens in Text-Rich Reasoning

    Oct 30, 2026Chanseul Cho, Noori Bae, Dongheon Lee +1VLM ReasoningVisually Grounded Reasoning

  16. RxnExBench: Benchmarking Vision-Language Models for Structured Chemical Reaction Extraction from Scientific Literature

    Oct 23, 2026Jiaxing Sun, Junyuan Gao, Ruikang Zhang +16Multimodal DatasetDocument Information Extraction

  17. ΨΨ-0.6: Adaptive Streams of Thought in a Unified Vision-Language World Model

    Oct 23, 2026Khai Loong Aw, Klemen Kotar, Baihan Zhang +5VLM ReasoningWorld Models for VLAs

  18. Separating Angular Alignment from Radial Ordering in Hyperbolic Vision-Language Models

    Oct 23, 2026Mingyang Gao, Yuwei Wu, Shiyu Luo +8Vision-Language ModelsHyperbolic Representation Learning

  19. OpenHome: From Perception to Insight in Smart-Home Camera Videos

    Oct 22, 2026Xianglong Wang, Neil Rodrigues, Pei Guo +9Video UnderstandingVLM Reasoning

  20. Where Did the Light Come From? Diagnosing Cast-Shadow Reasoning in VLMs and Borrowing the Missing Shadow Prior from Image Generation Models

    Oct 22, 2026Ryugo Morita, Shunsuke Kitada, Stanislav Frolov +5VLM ReasoningPhysical Reasoning

  21. OpenVLHarness: Real World Visual Agentic Tool-Use with Multimodal Memory

    Oct 19, 2026Yu Zhou, Cheng-Fu Yang, Rui Sun +8Tool Use in VLMsMultimodal Reasoning

  22. TempoMoE: Identifying and Steering Fast-Slow Thinking Experts for Efficient Multimodal Reasoning

    Oct 19, 2026Xiyue Zhang, Keru Zhou, Ziyi Wang +2Mixture-of-Experts InferenceEfficient Multimodal Inference

  23. Visual Scaffolding: Toward Native Multimodal Reasoning

    Oct 18, 2026Run Shao, Zhaoyang Zhang, Ziyu Li +6Multimodal ReasoningVLM Reasoning

  24. Video-Language Entailment for Video Similarity Estimation

    Oct 17, 2026Benedetta Liberatori, Lorenzo Vaquero, Elisa Ricci +1VLM Reasoning