Cover Image

Mitigating Over-Refusal in Aligned Large Language Models via Inference-Time Activation Energy

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Jiang, Eric Hanchen, Ou, Weixuan, Liu, Run, Pang, Shengyuan, Wan, Guancheng, Duan, Ranjie, Dong, Wei, Chang, Kai-Wei, Wang, XiaoFeng, Wu, Ying Nian, Li, Xinfeng
Format:	Preprint
Published:	2025
Subjects:	Machine Learning Artificial Intelligence Computation and Language
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

Dynamic Generation of Multi-LLM Agents Communication Topologies with Graph Diffusion Models
by: Jiang, Eric Hanchen, et al.
Published: (2025)

Learning to Rank Chain-of-Thought: Using a Small Model
by: Jiang, Eric Hanchen, et al.
Published: (2025)

Just Enough Shifts: Mitigating Over-Refusal in Aligned Language Models with Targeted Representation Fine-Tuning
by: Dabas, Mahavir, et al.
Published: (2025)

Over-Refusal and Representation Subspaces: A Mechanistic Analysis of Task-Conditioned Refusal in Aligned LLMs
by: Maskey, Utsav, et al.
Published: (2026)

Beyond Over-Refusal: Scenario-Based Diagnostics and Post-Hoc Mitigation for Exaggerated Refusals in LLMs
by: Yuan, Shuzhou, et al.
Published: (2025)

CompAlign: Improving Compositional Text-to-Image Generation with a Complex Benchmark and Fine-Grained Feedback
by: Wan, Yixin, et al.
Published: (2025)

The Male CEO and the Female Assistant: Evaluation and Mitigation of Gender Biases in Text-To-Image Generation of Dual Subjects
by: Wan, Yixin, et al.
Published: (2024)

White Men Lead, Black Women Help? Benchmarking and Mitigating Language Agency Social Biases in LLMs
by: Wan, Yixin, et al.
Published: (2024)

OR-Bench: An Over-Refusal Benchmark for Large Language Models
by: Cui, Justin, et al.
Published: (2024)

Improving Adversarial Robust Fairness via Anti-Bias Soft Label Distillation
by: Zhao, Shiji, et al.
Published: (2023)

Diagnose, Localize, Align: A Full-Stack Framework for Reliable LLM Multi-Agent Systems under Instruction Conflicts
by: Wan, Guancheng, et al.
Published: (2025)

Semantics-Adaptive Activation Intervention for LLMs via Dynamic Steering Vectors
by: Wang, Weixuan, et al.
Published: (2024)

NDM: A Noise-driven Detection and Mitigation Framework against Implicit Sexual Intentions in Text-to-Image Generation
by: Sun, Yitong, et al.
Published: (2025)

InsideOut: Measuring and Mitigating Insider-Outsider Bias in Interview Script Generation
by: Wan, Yixin, et al.
Published: (2025)

SafeConstellations: Mitigating Over-Refusals in LLMs Through Task-Aware Representation Steering
by: Maskey, Utsav, et al.
Published: (2025)

Oyster-I: Beyond Refusal -- Constructive Safety Alignment for Responsible Language Models
by: Duan, Ranjie, et al.
Published: (2025)

EVOREFUSE: Evolutionary Prompt Optimization for Evaluation and Mitigation of LLM Over-Refusal to Pseudo-Malicious Instructions
by: Wu, Xiaorui, et al.
Published: (2025)

Towards Class-wise Fair Adversarial Training via Anti-Bias Soft Label Distillation
by: Zhao, Shiji, et al.
Published: (2025)

Can LLMs Refuse Questions They Do Not Know? Measuring Knowledge-Aware Refusal in Factual Tasks
by: Pan, Wenbo, et al.
Published: (2025)

Unlocking the Potential of Text-to-Image Diffusion with PAC-Bayesian Theory
by: Jiang, Eric Hanchen, et al.
Published: (2024)

ORFuzz: Fuzzing the "Other Side" of LLM Safety -- Testing Over-Refusal
by: Zhang, Haonan, et al.
Published: (2025)

Mitigating Geospatial Knowledge Hallucination in Large Language Models: Benchmarking and Dynamic Factuality Aligning
by: Wang, Shengyuan, et al.
Published: (2025)

Refusal Before Decoding: Detecting and Exploiting Refusal Signals in Intermediate LLM Activations
by: Collu, Matteo Gioele, et al.
Published: (2026)

Stable Coresets via Posterior Sampling: Aligning Induced and Full Loss Landscapes
by: Chang, Wei-Kai, et al.
Published: (2025)

Programming Refusal with Conditional Activation Steering
by: Lee, Bruce W., et al.
Published: (2024)

Heuristic-Induced Multimodal Risk Distribution Jailbreak Attack for Multimodal Large Language Models
by: Teng, Ma, et al.
Published: (2024)

Think Before Refusal : Triggering Safety Reflection in LLMs to Mitigate False Refusal Behavior
by: Si, Shengyun, et al.
Published: (2025)

Please refuse to answer me! Mitigating Over-Refusal in Large Language Models via Adaptive Contrastive Decoding
by: Qi, Yupeng, et al.
Published: (2026)

FalseReject: A Resource for Improving Contextual Safety and Mitigating Over-Refusals in LLMs via Structured Reasoning
by: Zhang, Zhehao, et al.
Published: (2025)

Linearly Decoding Refused Knowledge in Aligned Language Models
by: Shrivastava, Aryan, et al.
Published: (2025)

Refusal Direction is Universal Across Safety-Aligned Languages
by: Wang, Xinpeng, et al.
Published: (2025)

SalesSim: Benchmarking and Aligning Multimodal Language Models as Retail User Simulators
by: Pruksachatkun, Yada, et al.
Published: (2026)

LatentRefusal: Latent-Signal Refusal for Unanswerable Text-to-SQL Queries
by: Ren, Xuancheng, et al.
Published: (2026)

Anchoring Refusal Direction: Mitigating Safety Risks in Tuning via Projection Constraint
by: Du, Yanrui, et al.
Published: (2025)

A Vision for Access Control in LLM-based Agent Systems
by: Li, Xinfeng, et al.
Published: (2025)

Adversarial Orthogonal Disentanglement for LVLM Hallucination Mitigation
by: Cheng, Ruoxi, et al.
Published: (2026)

Beyond No: Quantifying AI Over-Refusal and Emotional Attachment Boundaries
by: Noever, David, et al.
Published: (2025)

Legilimens: Practical and Unified Content Moderation for Large Language Model Services
by: Wu, Jialin, et al.
Published: (2024)

Learning to Refuse: Towards Mitigating Privacy Risks in LLMs
by: Liu, Zhenhua, et al.
Published: (2024)

Tracing the Dynamics of Refusal: Exploiting Latent Refusal Trajectories for Robust Jailbreak Detection
by: Hu, Xulin, et al.
Published: (2026)