activity
20222025
most citedSleeper Agents: Training Deceptive LLMs that Persist Through Safety Training

39 citations · 155 across the 12 of their papers we have counts for

collaborators
Showing cs.AIShow all

5 papers · 1 filter

cs.AI2025

Natural Emergent Misalignment from Reward Hacking in Production RL

Monte MacDiarmid, Benjamin Wright, Jonathan Uesato +19

We show that when large language models learn to reward hack on production RL environments, this can result in egregious emergent misalignment. We start with a pretrained model, im…

cs.AI2025

Chain of Thought Monitorability: A New and Fragile Opportunity for AI Safety

Tomek Korbak, Mikita Balesni, Elizabeth Barnes +38

AI systems that "think" in human language offer a unique opportunity for AI safety: we can monitor their chains of thought (CoT) for the intent to misbehave. Like all other known A…

cs.AI20253 cited

Auditing language models for hidden objectives

Samuel Marks, Johannes Treutlein, Trenton Bricken +32

We study the feasibility of conducting alignment audits: investigations into whether models have undesired objectives. As a testbed, we train a language model with a hidden objecti…

cs.AI202424 cited

Alignment faking in large language models

Ryan Greenblatt, Carson Denison, Benjamin Wright +17

We present a demonstration of a large language model engaging in alignment faking: selectively complying with its training objective in training to prevent modification of its beha…

cs.AI20249 cited

Sycophancy to Subterfuge: Investigating Reward-Tampering in Large Language Models

Carson Denison, Monte MacDiarmid, Fazl Barez +11

In reinforcement learning, specification gaming occurs when AI systems learn undesired behaviors that are highly rewarded due to misspecified training goals. Specification gaming c…