7 papers
An Independent Safety Evaluation of Kimi K2.5
Zheng-Xin Yong, Parv Mahajan, Andy Wang +12
Kimi K2.5 is an open-weight LLM that rivals closed models across coding, multimodal, and agentic benchmarks, but was released without an accompanying safety evaluation. In this wor…
The Persistent Vulnerability of Aligned AI Systems
Aengus Lynch
Autonomous AI agents are being deployed with filesystem access, email control, and multi-step planning. This thesis contributes to four open problems in AI safety: understanding da…
Agentic Misalignment: How LLMs Could Be Insider Threats
Aengus Lynch, Benjamin Wright, Caleb Larson +5
We stress-tested 16 leading models from multiple developers in hypothetical corporate environments to identify potentially risky agentic behaviors before they cause real harm. In t…
Latent Adversarial Training Improves Robustness to Persistent Harmful Behaviors in LLMs
Abhay Sheshadri, Aidan Ewart, Phillip Guo +8
Large language models (LLMs) can often be made to behave in undesirable ways that they are explicitly fine-tuned not to. For example, the LLM red-teaming literature has produced a…
Analyzing the Generalization and Reliability of Steering Vectors
Daniel Tan, David Chanin, Aengus Lynch +4
Steering vectors (SVs) have been proposed as an effective approach to adjust language model behaviour at inference time by intervening on intermediate model activations. They have…
How Do Large Language Monkeys Get Their Power (Laws)?
Rylan Schaeffer, Joshua Kazdan, John Hughes +7
Recent research across mathematical problem solving, proof assistant programming and multimodal jailbreaking documents a striking finding: when (multimodal) language model tackle a…