AI Safety

The Automated but Risky Game: Modeling Agent-to-Agent Negotiations and Transactions in Consumer Markets EMNLP 2025 The Canary’s Echo: Auditing Privacy Risks of LLM-Generated Synthetic Text ICML 2025 The Confidence Paradox: Can LLM Know When It’s Wrong? IJCNLP 2025 The Dangers of Indirect Prompt Injection Attacks on LLM-based Autonomous Web Navigation Agents: A Demonstration EMNLP 2025 The Dark Side of Function Calling: Pathways to Jailbreaking Large Language Models COLING 2025 The Elephant in the Room: Exploring the Role of Neutral Words in Language Model Group-Agnostic Debiasing ACL 2025 The Elicitation Game: Evaluating Capability Elicitation Techniques ICML 2025 The Energy Loss Phenomenon in RLHF: A New Perspective on Mitigating Reward Hacking ICML 2025 The Geometry of Refusal in Large Language Models: Concept Cones and Representational Independence ICML 2025 The Great Contradiction Showdown: How Jailbreak and Stealth Wrestle in Vision-Language Models? ACML 2025 The Hallucination Tax of Reinforcement Finetuning EMNLP 2025 The Hidden Dimensions of LLM Alignment: A Multi-Dimensional Analysis of Orthogonal Safety Directions ICML 2025 The Hidden Risks of Large Reasoning Models: A Safety Assessment of R1 IJCNLP 2025 The Impact of Negated Text on Hallucination with Large Language Models EMNLP 2025 The Jailbreak Tax: How Useful are Your Jailbreak Outputs? ICML 2025 The Law of Knowledge Overshadowing: Towards Understanding, Predicting and Preventing LLM Hallucination ACL 2025 The Law of Knowledge Overshadowing: Towards Understanding, Predicting, and Preventing LLM Hallucination ACL 2025 The Lawyer That Never Thinks: Consistency and Fairness as Keys to Reliable AI ACL 2025 The Limits of Predicting Agents from Behaviour ICML 2025 The Only Way is Ethics: A Guide to Ethical Research with Large Language Models COLING 2025 The Partially Observable Off-Switch Game AAAI 2025 The Perils of Optimizing Learned Reward Functions: Low Training Error Does Not Guarantee Low Regret ICML 2025 The RAG Paradox: A Black-Box Attack Exploiting Unintentional Vulnerabilities in Retrieval-Augmented Generation Systems EMNLP 2025 The Rise of Darkness: Safety-Utility Trade-Offs in Role-Playing Dialogue Agents ACL 2025 The Security Threat of Compressed Projectors in Large Vision-Language Models EMNLP 2025