Cover Image

Not All Tokens See Equally: Perception-Grounded Policy Optimization for Large Vision-Language Models

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Ye, Zekai, Li, Qiming, Feng, Xiaocheng, Chen, Ruihan, Li, Ziming, Ren, Haoyu, Chen, Kun, Tu, Dandan, Qin, Bing
Format:	Preprint
Published:	2026
Subjects:	Artificial Intelligence
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

MPR-GUI: Benchmarking and Enhancing Multilingual Perception and Reasoning in GUI Agents
by: Chen, Ruihan, et al.
Published: (2025)

Unlocking Multilingual Reasoning Capability of LLMs and LVLMs through Representation Engineering
by: Li, Qiming, et al.
Published: (2025)

CAI: Caption-Sensitive Attention Intervention for Mitigating Object Hallucination in Large Vision-Language Models
by: Li, Qiming, et al.
Published: (2025)

CAST: Mitigating Object Hallucination in Large Vision-Language Models via Caption-Guided Visual Attention Steering
by: Li, Qiming, et al.
Published: (2026)

Causal Tracing of Object Representations in Large Vision Language Models: Mechanistic Interpretability and Hallucination Mitigation
by: Li, Qiming, et al.
Published: (2025)

CLAIM: Mitigating Multilingual Object Hallucination in Large Vision-Language Models with Cross-Lingual Attention Intervention
by: Ye, Zekai, et al.
Published: (2025)

Culture-Aware Machine Translation in Large Language Models: Benchmarking and Investigation
by: Yuan, Zekun, et al.
Published: (2026)

Not All Tokens Matter Equally: Dynamic In-context Vector Distillation with Decisive-Token Supervision for Long-form Medical Report Generation
by: Wu, Ning, et al.
Published: (2026)

Not All Tokens and Heads Are Equally Important: Dual-Level Attention Intervention for Hallucination Mitigation
by: Tang, Lexiang, et al.
Published: (2025)

Learning Fine-Grained Grounded Citations for Attributed Large Language Models
by: Huang, Lei, et al.
Published: (2024)

Investigating and Mitigating the Multimodal Hallucination Snowballing in Large Vision-Language Models
by: Zhong, Weihong, et al.
Published: (2024)

Not All Qubits are Utilized Equally
by: Pope, Jeremie, et al.
Published: (2025)

Not All Alumni Engage Equally
Published: (2024)

Not All Data Are Unlearned Equally
by: Krishnan, Aravind, et al.
Published: (2025)

Cross-Lingual Text-Rich Visual Comprehension: An Information Theory Perspective
by: Yu, Xinmiao, et al.
Published: (2024)

Not All Noises Are Created Equally:Diffusion Noise Selection and Optimization
by: Qi, Zipeng, et al.
Published: (2024)

CultureForest: Understanding and Evaluating Cultural Norm Grounded Reasoning in LLMs
by: Ye, Yangfan, et al.
Published: (2026)

Efficient NeRF Optimization -- Not All Samples Remain Equally Hard
by: Korhonen, Juuso, et al.
Published: (2024)

LangGPS: Language Separability Guided Data Pre-Selection for Joint Multilingual Instruction Tuning
by: Ye, Yangfan, et al.
Published: (2025)

Jigsaw-Puzzles: From Seeing to Understanding to Reasoning in Vision-Language Models
by: Lyu, Zesen, et al.
Published: (2025)

Not Every Rubric Teaches Equally: Policy-Aware Rubric Rewards for RLVR
by: Tyagi, Utkarsh, et al.
Published: (2026)

x1: Learning to Think Adaptively Across Languages and Cultures
by: Ye, Yangfan, et al.
Published: (2026)

Not All Explanations for Deep Learning Phenomena Are Equally Valuable
by: Jeffares, Alan, et al.
Published: (2025)

Not All Samples Should Be Utilized Equally: Towards Understanding and Improving Dataset Distillation
by: Wang, Shaobo, et al.
Published: (2024)

Stratagem: Learning Transferable Reasoning via Trajectory-Modulated Game Self-Play
by: Feng, Xiachong, et al.
Published: (2026)

ROVI: A VLM-LLM Re-Captioned Dataset for Open-Vocabulary Instance-Grounded Text-to-Image Generation
by: Peng, Cihang, et al.
Published: (2025)

Enhancing Non-English Capabilities of English-Centric Large Language Models through Deep Supervision Fine-Tuning
by: Huo, Wenshuai, et al.
Published: (2025)

Balanced Token Pruning: Accelerating Vision Language Models Beyond Local Optimization
by: Li, Kaiyuan, et al.
Published: (2025)

Context-Aware Hierarchical Taxonomy Generation for Scientific Papers via LLM-Guided Multi-Aspect Clustering
by: Zhu, Kun, et al.
Published: (2025)

NCT-CRC-HE: Not All Histopathological Datasets Are Equally Useful
by: Ignatov, Andrey, et al.
Published: (2024)

Act to See, See to Act: Diffusion-Driven Perception-Action Interplay for Adaptive Policies
by: Wang, Jing, et al.
Published: (2025)

From Seeing to Thinking: Decoupling Perception and Reasoning Improves Post-Training of Vision-Language Models
by: Wu, Juncheng, et al.
Published: (2026)

SeeGround: See and Ground for Zero-Shot Open-Vocabulary 3D Visual Grounding
by: Li, Rong, et al.
Published: (2024)

Visual-Instructed Degradation Diffusion for All-in-One Image Restoration
by: Luo, Wenyang, et al.
Published: (2025)

SeeNav-Agent: Enhancing Vision-Language Navigation with Visual Prompt and Step-Level Policy Optimization
by: Wang, Zhengcheng, et al.
Published: (2025)

Not All Instances Are Equally Valuable: Towards Influence-Weighted Dataset Distillation
by: Deng, Qiyan, et al.
Published: (2025)

Not All Factors Crowd Equally: Modeling, Measuring, and Trading on Alpha Decay
by: Lee, Chorok
Published: (2025)

One for All: Update Parameterized Knowledge Across Multiple Models
by: Ma, Weitao, et al.
Published: (2025)

Not All Tasks Are Equally Difficult: Multi-Task Deep Reinforcement Learning with Dynamic Depth Routing
by: He, Jinmin, et al.
Published: (2023)

All-in-One: Transferring Vision Foundation Models into Stereo Matching
by: Zhou, Jingyi, et al.
Published: (2024)