Image de couverture de livre

Compromising Honesty and Harmlessness in Language Models via Deception Attacks

Fuente: arXiv

Enregistré dans:

Détails bibliographiques
Auteurs principaux:	Vaugrante, Laurène, Carlon, Francesca, Menke, Maluna, Hagendorff, Thilo
Format:	Preprint
Publié:	2025
Sujets:	Computation and Language Artificial Intelligence 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

PRIDE -- Parameter-Efficient Reduction of Identity Discrimination for Equality in LLMs
par: Menke, Maluna, et autres
Publié: (2025)

A Looming Replication Crisis in Evaluating Behavior in Language Models? Evidence and Solutions
par: Vaugrante, Laurène, et autres
Publié: (2024)

Emergently Misaligned Language Models Show Behavioral Self-Awareness That Shifts With Subsequent Realignment
par: Vaugrante, Laurène, et autres
Publié: (2026)

Deception Abilities Emerged in Large Language Models
par: Hagendorff, Thilo
Publié: (2023)

On the Inevitability of Left-Leaning Political Bias in Aligned Language Models
par: Hagendorff, Thilo
Publié: (2025)

Mapping the Ethics of Generative AI: A Comprehensive Scoping Review
par: Hagendorff, Thilo
Publié: (2024)

Fairness Hacking: The Malicious Practice of Shrouding Unfairness in Algorithms
par: Meding, Kristof, et autres
Publié: (2023)

Evaluation Awareness in Language Models Has Limited Effect on Behaviour
par: Knecht, Amelie, et autres
Publié: (2026)

Mix Data or Merge Models? Balancing the Helpfulness, Honesty, and Harmlessness of Large Language Model via Model Merging
par: Yang, Jinluan, et autres
Publié: (2025)

Unmasking and Improving Data Credibility: A Study with Datasets for Training Harmless Language Models
par: Zhu, Zhaowei, et autres
Publié: (2023)

Large Reasoning Models Are Autonomous Jailbreak Agents
par: Hagendorff, Thilo, et autres
Publié: (2025)

Speciesism in AI: Evaluating Discrimination Against Animals in Large Language Models
par: Jotautaitė, Monika, et autres
Publié: (2025)

The MASK Benchmark: Disentangling Honesty From Accuracy in AI Systems
par: Ren, Richard, et autres
Publié: (2025)

A Survey on the Honesty of Large Language Models
par: Li, Siheng, et autres
Publié: (2024)

Why Agents Compromise Safety Under Pressure
par: Jiang, Hengle, et autres
Publié: (2026)

BeHonest: Benchmarking Honesty in Large Language Models
par: Chern, Steffi, et autres
Publié: (2024)

Alignment for Honesty
par: Yang, Yuqing, et autres
Publié: (2023)

"Dark Triad" Model Organisms of Misalignment: Narrow Fine-Tuning Mirrors Human Antisocial Behavior
par: Lulla, Roshni, et autres
Publié: (2026)

Attacks on Third-Party APIs of Large Language Models
par: Zhao, Wanru, et autres
Publié: (2024)

When Image Generation Goes Wrong: A Safety Analysis of Stable Diffusion Models
par: Schneider, Matthias, et autres
Publié: (2024)

Beyond Chains of Thought: Benchmarking Latent-Space Reasoning Abilities in Large Language Models
par: Hagendorff, Thilo, et autres
Publié: (2025)

Training-Free Cultural Alignment of Large Language Models via Persona Disagreement
par: Kiet, Huynh Trung, et autres
Publié: (2026)

Raising the Bar: Investigating the Values of Large Language Models via Generative Evolving Testing
par: Jiang, Han, et autres
Publié: (2024)

Self-Alignment of Large Language Models via Monopolylogue-based Social Scene Simulation
par: Pang, Xianghe, et autres
Publié: (2024)

Denevil: Towards Deciphering and Navigating the Ethical Values of Large Language Models via Instruction Learning
par: Duan, Shitong, et autres
Publié: (2023)

IROTE: Human-like Traits Elicitation of Large Language Model via In-Context Self-Reflective Optimization
par: Bai, Yuzhuo, et autres
Publié: (2025)

Cross-Language Bias Examination in Large Language Models
par: Liang, Yuxuan, et autres
Publié: (2025)

Smart Trial: Evaluating the Use of Large Language Models for Recruiting Clinical Trial Participants via Social Media
par: Zhou, Xiaofan, et autres
Publié: (2025)

On the Creativity of Large Language Models
par: Franceschelli, Giorgio, et autres
Publié: (2023)

Evaluating Large Language Models for Detecting Antisemitism
par: Patel, Jay, et autres
Publié: (2025)

How Large Language Models are Designed to Hallucinate
par: Ackermann, Richard, et autres
Publié: (2025)

The Polite Liar: Epistemic Pathology in Language Models
par: DeVilling, Bentley
Publié: (2025)

Anticipating Innovation Using Large Language Models
par: Fenoaltea, Enrico Maria, et autres
Publié: (2026)

Open-Ended Wargames with Large Language Models
par: Hogan, Daniel P., et autres
Publié: (2024)

Large Language Models for Education: A Survey
par: Xu, Hanyi, et autres
Publié: (2024)

Large Language Models as Misleading Assistants in Conversation
par: Hou, Betty Li, et autres
Publié: (2024)

Evaluating Psychological Safety of Large Language Models
par: Li, Xingxuan, et autres
Publié: (2022)

Large Language Models for Medicine: A Survey
par: Zheng, Yanxin, et autres
Publié: (2024)

Are Large Language Models Good Essay Graders?
par: Kundu, Anindita, et autres
Publié: (2024)

All in How You Ask for It: Simple Black-Box Method for Jailbreak Attacks
par: Takemoto, Kazuhiro
Publié: (2024)