4 accepted papers
We partially reverse-engineer a convolutional recurrent neural network (RNN) trained with model-free reinforcement learning to play the box-pushing game Sokoban. We find that the RNN stores future moves (plans) as activations in particular channels of the hidden state, which we call *path channels*.…
Frontier AI developers are relying on layers of safeguards to protect against catastrophic misuse of AI systems. Anthropic guards their latest Claude 4 Opus model using one such defense pipeline, and other frontier developers including Google DeepMind and OpenAI pledge to soon deploy similar defense
Increasing model size has unlocked a dazzling array of capabilities in language models. At the same time, even frontier models remain vulnerable to jailbreaks and prompt injections, despite concerted efforts to make them robust. As both attackers and defenders gain access to more compute, and as mod…
We propose coactive learning as a model and feedback mechanism for training large language models (LLMs). The key insight is that users provide implicit feedback whenever they edit the text $y$ proposed by an LLM. While the edited text $\bar y$ is typically not a gold-standard example for supervised…