Cover Image

Think-Reflect-Revise: A Policy-Guided Reflective Framework for Safety Alignment in Large Vision Language Models

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Weng, Fenghua, Lu, Chaochao, Hu, Xia, Shao, Wenqi, Wang, Wenjie
Format:	Preprint
Published:	2025
Subjects:	Computer Vision and Pattern Recognition Computation and Language
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

Look Again, Think Slowly: Enhancing Visual Reflection in Vision-Language Models
by: Jian, Pu, et al.
Published: (2025)

MSR-Align: Policy-Grounded Multimodal Alignment for Safety-Aware Reasoning in Vision-Language Models
by: Xia, Yinan, et al.
Published: (2025)

Large Vision-Language Model Alignment and Misalignment: A Survey Through the Lens of Explainability
by: Shu, Dong, et al.
Published: (2025)

Quantifying and Mitigating Unimodal Biases in Multimodal Large Language Models: A Causal Perspective
by: Chen, Meiqi, et al.
Published: (2024)

SPA-VL: A Comprehensive Safety Preference Alignment Dataset for Vision Language Model
by: Zhang, Yongting, et al.
Published: (2024)

VEGAS: Mitigating Hallucinations in Large Vision-Language Models via Vision-Encoder Attention Guided Adaptive Steering
by: Wang, Zihu, et al.
Published: (2025)

Layer-wise Alignment: Examining Safety Alignment Across Image Encoder Layers in Vision Language Models
by: Bachu, Saketh, et al.
Published: (2024)

Evolving Contextual Safety in Multi-Modal Large Language Models via Inference-Time Self-Reflective Memory
by: Zhang, Ce, et al.
Published: (2026)

Prompt-Induced Score Variance in Zero-Shot Binary Vision-Language Safety Classification
by: Weng, Charles, et al.
Published: (2026)

URPO: A Unified Reward & Policy Optimization Framework for Large Language Models
by: Lu, Songshuo, et al.
Published: (2025)

DREAM: Disentangling Risks to Enhance Safety Alignment in Multimodal Large Language Models
by: Liu, Jianyu, et al.
Published: (2025)

AlignMMBench: Evaluating Chinese Multimodal Alignment in Large Vision-Language Models
by: Wu, Yuhang, et al.
Published: (2024)

EgoThink: Evaluating First-Person Perspective Thinking Capability of Vision-Language Models
by: Cheng, Sijie, et al.
Published: (2023)

VLFeedback: A Large-Scale AI Feedback Dataset for Large Vision-Language Models Alignment
by: Li, Lei, et al.
Published: (2024)

Rethinking Bottlenecks in Safety Fine-Tuning of Vision Language Models
by: Ding, Yi, et al.
Published: (2025)

Vision-Language Models Can Self-Improve Reasoning via Reflection
by: Cheng, Kanzhi, et al.
Published: (2024)

Fast-Slow Thinking GRPO for Large Vision-Language Model Reasoning
by: Xiao, Wenyi, et al.
Published: (2025)

MIRL: Mutual Information-Guided Reinforcement Learning for Vision-Language Models
by: Zhang, Yin, et al.
Published: (2026)

Deciphering Cross-Modal Alignment in Large Vision-Language Models with Modality Integration Rate
by: Huang, Qidong, et al.
Published: (2024)

How Does Vision-Language Adaptation Impact the Safety of Vision Language Models?
by: Lee, Seongyun, et al.
Published: (2024)

RefLSM: Linearized Structural-Prior Reflectance Model for Medical Image Segmentation and Bias-Field Correction
by: Zhao, Wenqi, et al.
Published: (2025)

Mitigating Hallucinations in Large Vision-Language Models by Self-Injecting Hallucinations
by: Lu, Yifan, et al.
Published: (2025)

Integrating Chain-of-Thought for Multimodal Alignment: A Study on 3D Vision-Language Learning
by: Chen, Yanjun, 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)

Reflective Instruction Tuning: Mitigating Hallucinations in Large Vision-Language Models
by: Zhang, Jinrui, et al.
Published: (2024)

Re-Thinking Inverse Graphics With Large Language Models
by: Kulits, Peter, et al.
Published: (2024)

Hierarchical Safety Realignment: Lightweight Restoration of Safety in Pruned Large Vision-Language Models
by: Li, Yue, et al.
Published: (2025)

Enhancing Large Vision Language Models with Self-Training on Image Comprehension
by: Deng, Yihe, et al.
Published: (2024)

A Unified Hallucination Mitigation Framework for Large Vision-Language Models
by: Chang, Yue, et al.
Published: (2024)

HiMo-CLIP: Modeling Semantic Hierarchy and Monotonicity in Vision-Language Alignment
by: Wu, Ruijia, et al.
Published: (2025)

ETHER: Efficient Finetuning of Large-Scale Models with Hyperplane Reflections
by: Bini, Massimo, et al.
Published: (2024)

MARS: Memory-Enhanced Agents with Reflective Self-improvement
by: Liang, Xuechen, 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)

CELLO: Causal Evaluation of Large Vision-Language Models
by: Chen, Meiqi, et al.
Published: (2024)

Seeing No Evil: Blinding Large Vision-Language Models to Safety Instructions via Adversarial Attention Hijacking
by: Li, Jingru, et al.
Published: (2026)

Cross-Modal Safety Mechanism Transfer in Large Vision-Language Models
by: Xu, Shicheng, et al.
Published: (2024)

Conflict Adaptation in Vision-Language Models
by: Hu, Xiaoyang
Published: (2025)

X-Reflect: Cross-Reflection Prompting for Multimodal Recommendation
by: Lyu, Hanjia, et al.
Published: (2024)

PUMGPT: A Large Vision-Language Model for Product Understanding
by: Xue, Wei, et al.
Published: (2023)

Evaluation and Enhancement of Semantic Grounding in Large Vision-Language Models
by: Lu, Jiaying, et al.
Published: (2023)