Cover Image

More Thought, Less Accuracy? On the Dual Nature of Reasoning in Vision-Language Models

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Tian, Xinyu, Zou, Shu, Yang, Zhaoyuan, He, Mengqi, Waschkowski, Fabian, Wesemann, Lukas, Tu, Peter, Zhang, Jing
Format:	Preprint
Published:	2025
Subjects:	Computer Vision and Pattern Recognition Artificial Intelligence
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

Unlocking Vision-Language Models for Video Anomaly Detection via Fine-Grained Prompting
by: Zou, Shu, et al.
Published: (2025)

All Roads Lead to Rome: Incentivizing Divergent Thinking in Vision-Language Models
by: Tian, Xinyu, et al.
Published: (2026)

Identifying and Mitigating Position Bias of Multi-image Vision-Language Models
by: Tian, Xinyu, et al.
Published: (2025)

ArGue: Attribute-Guided Prompt Tuning for Vision-Language Models
by: Tian, Xinyu, et al.
Published: (2023)

High-Entropy Tokens as Multimodal Failure Points in Vision-Language Models
by: He, Mengqi, et al.
Published: (2025)

Black Sheep in the Herd: Playing with Spuriously Correlated Attributes for Vision-Language Recognition
by: Tian, Xinyu, et al.
Published: (2025)

SimLabel: Consistency-Guided OOD Detection with Pretrained Vision-Language Models
by: Zou, Shu, et al.
Published: (2025)

Robust Spatiotemporal Forecasting Using Adaptive Deep-Unfolded Variational Mode Decomposition
by: Ahmad, Osama, et al.
Published: (2025)

Variational Mode-Driven Graph Convolutional Network for Spatiotemporal Traffic Forecasting
by: Ahmad, Osama, et al.
Published: (2024)

Break the Brake, Not the Wheel: Untargeted Jailbreak via Entropy Maximization
by: He, Mengqi, et al.
Published: (2026)

Less-to-More Generalization: Unlocking More Controllability by In-Context Generation
by: Wu, Shaojin, et al.
Published: (2025)

Less is More: Lean yet Powerful Vision-Language Model for Autonomous Driving
by: Yang, Sheng, et al.
Published: (2025)

DualCoT-VLA: Visual-Linguistic Chain of Thought via Parallel Reasoning for Vision-Language-Action Models
by: Zhong, Zhide, et al.
Published: (2026)

FantasyVLN: Unified Multimodal Chain-of-Thought Reasoning for Vision-Language Navigation
by: Zuo, Jing, et al.
Published: (2026)

GTR-Bench: Evaluating Geo-Temporal Reasoning in Vision-Language Models
by: Xie, Qinghongbing, et al.
Published: (2025)

More Thinking, Less Seeing? Assessing Amplified Hallucination in Multimodal Reasoning Models
by: Liu, Chengzhi, et al.
Published: (2025)

Select Less, Reason More: Prioritizing Evidence Purity for Video Reasoning
by: Li, Xuchen, et al.
Published: (2025)

Look Less, Reason More: Rollout-Guided Adaptive Pixel-Space Reasoning
by: Li, Xuchen, et al.
Published: (2025)

Measuring and Improving Chain-of-Thought Reasoning in Vision-Language Models
by: Chen, Yangyi, et al.
Published: (2023)

ReasoningTrack: Chain-of-Thought Reasoning for Long-term Vision-Language Tracking
by: Wang, Xiao, et al.
Published: (2025)

DreamSteerer: Enhancing Source Image Conditioned Editability using Personalized Diffusion Models
by: Yu, Zhengyang, et al.
Published: (2024)

LIME: Less Is More for MLLM Evaluation
by: Zhu, King, et al.
Published: (2024)

Multimodal Chain of Continuous Thought for Latent-Space Reasoning in Vision-Language Models
by: Pham, Tan-Hanh, et al.
Published: (2025)

C-CoT: Counterfactual Chain-of-Thought with Vision-Language Models for Safe Autonomous Driving
by: Tian, Kefei, et al.
Published: (2026)

NuScenes-SpatialQA: A Spatial Understanding and Reasoning Benchmark for Vision-Language Models in Autonomous Driving
by: Tian, Kexin, et al.
Published: (2025)

DualReal: Adaptive Joint Training for Lossless Identity-Motion Fusion in Video Customization
by: Wang, Wenchuan, et al.
Published: (2025)

Less Precise Can Be More Reliable: A Systematic Evaluation of Quantization's Impact on VLMs Beyond Accuracy
by: Bouguerra, Aymen, et al.
Published: (2025)

LPT: Less-overfitting Prompt Tuning for Vision-Language Model
by: Ding, Chenhao, et al.
Published: (2024)

AIM-CoT: Active Information-driven Multimodal Chain-of-Thought for Vision-Language Reasoning
by: Li, Xiping, et al.
Published: (2025)

Think-as-You-See: Streaming Chain-of-Thought Reasoning for Large Vision-Language Models
by: Zhang, Jialiang, et al.
Published: (2026)

IMPUS: Image Morphing with Perceptually-Uniform Sampling Using Diffusion Models
by: Yang, Zhaoyuan, et al.
Published: (2023)

Argus: Vision-Centric Reasoning with Grounded Chain-of-Thought
by: Man, Yunze, et al.
Published: (2025)

Learning More by Seeing Less: Structure First Learning for Efficient, Transferable, and Human-Aligned Vision
by: Li, Tianqin, et al.
Published: (2025)

Towards Faithful Reasoning in Remote Sensing: A Perceptually-Grounded GeoSpatial Chain-of-Thought for Vision-Language Models
by: Liu, Jiaqi, et al.
Published: (2025)

TinySAM 2: Extreme Memory Compression for Efficient Track Anything Model
by: Ding, Zhaoyuan, et al.
Published: (2026)

The Dual Mechanisms of Spatial Reasoning in Vision-Language Models
by: Cui, Kelly, et al.
Published: (2026)

Less is More: Discovering Concise Network Explanations
by: Kondapaneni, Neehar, et al.
Published: (2024)

Probability Density Geodesics in Image Diffusion Latent Space
by: Yu, Qingtao, et al.
Published: (2025)

AgentThink: A Unified Framework for Tool-Augmented Chain-of-Thought Reasoning in Vision-Language Models for Autonomous Driving
by: Qian, Kangan, et al.
Published: (2025)

Boosting the Generalization and Reasoning of Vision Language Models with Curriculum Reinforcement Learning
by: Deng, Huilin, et al.
Published: (2025)