Image de couverture de livre

Auditing Agent Harness Safety

Fuente: arXiv

Enregistré dans:

Détails bibliographiques
Auteurs principaux:	Liu, Chengzhi, Guo, Yichen, Liu, Yepeng, Yang, Yuzhe, Yan, Qianqi, Zhao, Xuandong, Hua, Wenyue, Liu, Sheng, Li, Sharon, Bu, Yuheng, Wang, Xin Eric
Format:	Preprint
Publié:	2026
Sujets:	Computation and Language Computers and Society
Accès en ligne:	Acceder al recurso
Tags:	Ajouter un tag Pas de tags, Soyez le premier à ajouter un tag!

Documents similaires

Dataset Protection via Watermarked Canaries in Retrieval-Augmented LLMs
par: Liu, Yepeng, et autres
Publié: (2025)

In-Context Watermarks for Large Language Models
par: Liu, Yepeng, et autres
Publié: (2025)

Adaptive Text Watermark for Large Language Models
par: Liu, Yepeng, et autres
Publié: (2024)

Position: LLM Watermarking Should Align Stakeholders' Incentives for Practical Adoption
par: Liu, Yepeng, et autres
Publié: (2025)

WorldMemArena: Evaluating Multimodal Agent Memory Through Action-World Interaction
par: Liu, Chengzhi, et autres
Publié: (2026)

Multimodal Situational Safety
par: Zhou, Kaiwen, et autres
Publié: (2024)

The Hidden Risks of Large Reasoning Models: A Safety Assessment of R1
par: Zhou, Kaiwen, et autres
Publié: (2025)

Beyond the Safety Bundle: Auditing the Helpful and Harmless Dataset
par: Chehbouni, Khaoula, et autres
Publié: (2024)

Reasoning Within the Mind: Dynamic Multimodal Interleaving in Latent Space
par: Liu, Chengzhi, et autres
Publié: (2025)

War and Peace (WarAgent): Large Language Model-based Multi-Agent Simulation of World Wars
par: Hua, Wenyue, et autres
Publié: (2023)

The Silicon Ceiling: Auditing GPT's Race and Gender Biases in Hiring
par: Armstrong, Lena, et autres
Publié: (2024)

Defending LLM Watermarking Against Spoofing Attacks with Contrastive Representation Learning
par: An, Li, et autres
Publié: (2025)

Exploring Novelty Differences between Industry and Academia: A Knowledge Entity-centric Perspective
par: Zhao, Hongye, et autres
Publié: (2026)

A Multimodal, Multilingual, and Multidimensional Pipeline for Fine-grained Crowdsourcing Earthquake Damage Evaluation
par: Ma, Zihui, et autres
Publié: (2025)

The Better Angels of Machine Personality: How Personality Relates to LLM Safety
par: Zhang, Jie, et autres
Publié: (2024)

Length Value Model: Scalable Value Pretraining for Token-Level Length Modeling
par: Zhang, Zhen, et autres
Publié: (2026)

Assessing Judging Bias in Large Reasoning Models: An Empirical Study
par: Wang, Qian, et autres
Publié: (2025)

Analyzing the Safety of Japanese Large Language Models in Stereotype-Triggering Prompts
par: Nakanishi, Akito, et autres
Publié: (2025)

From Attribution to Abstention: Training-Free Attention-Based Auditing for Clinical Summarization
par: Yan, Qianqi, et autres
Publié: (2026)

Auditing Stance Asymmetry in Generative Explanations
par: Han, Jiarui
Publié: (2026)

Presenting a Paper is an Art: Self-Improvement Aesthetic Agents for Academic Presentations
par: Liu, Chengzhi, et autres
Publié: (2025)

AuditWen:An Open-Source Large Language Model for Audit
par: Huang, Jiajia, et autres
Publié: (2024)

Exploring Gender Biases in Language Patterns of Human-Conversational Agent Conversations
par: Liu, Weizi
Publié: (2024)

Guardians and Offenders: A Survey on Harmful Content Generation and Safety Mitigation of LLM
par: Zhang, Chi, et autres
Publié: (2025)

ConVerse: Benchmarking Contextual Safety in Agent-to-Agent Conversations
par: Gomaa, Amr, et autres
Publié: (2025)

EmpathyAgent: Can Embodied Agents Conduct Empathetic Actions?
par: Chen, Xinyan, et autres
Publié: (2025)

REALM: A Dataset of Real-World LLM Use Cases
par: Cheng, Jingwen, et autres
Publié: (2025)

Persuasion Dynamics in LLMs: Investigating Robustness and Adaptability in Knowledge and Safety with DuET-PD
par: Tan, Bryan Chen Zhengyu, et autres
Publié: (2025)

Safety Evaluation and Enhancement of DeepSeek Models in Chinese Contexts
par: Zhang, Wenjing, et autres
Publié: (2025)

Characterizing Selective Refusal Bias in Large Language Models
par: Khorramrouz, Adel, et autres
Publié: (2025)

EcoLANG: Efficient and Effective Agent Communication Language Induction for Social Simulation
par: Mou, Xinyi, et autres
Publié: (2025)

AgentSense: Benchmarking Social Intelligence of Language Agents through Interactive Scenarios
par: Mou, Xinyi, et autres
Publié: (2024)

A Comparative Evaluation of Structural Topic Models and BERTopic for Short, Open-Ended Survey Responses
par: Jiang, Yan, et autres
Publié: (2026)

Audit Me If You Can: Query-Efficient Active Fairness Auditing of Black-Box LLMs
par: Hartmann, David, et autres
Publié: (2026)

The Staircase of Ethics: Probing LLM Value Priorities through Multi-Step Induction to Complex Moral Dilemmas
par: Wu, Ya, et autres
Publié: (2025)

OmniTrace: A Unified Framework for Generation-Time Attribution in Omni-Modal LLMs
par: Yan, Qianqi, et autres
Publié: (2026)

SOTOPIA-$Ω$: Dynamic Strategy Injection Learning and Social Instruction Following Evaluation for Social Agents
par: Zhang, Wenyuan, et autres
Publié: (2025)

On the Suitability of LLM-Driven Agents for Dark Pattern Audits
par: Sun, Chen, et autres
Publié: (2026)

Aligned but Blind: Alignment Increases Implicit Bias by Reducing Awareness of Race
par: Sun, Lihao, et autres
Publié: (2025)

How Did We Get Here? Summarizing Conversation Dynamics
par: Hua, Yilun, et autres
Publié: (2024)