Private Event
Reading Group & Discussion: Reasoning Models Struggle to Control their Chains of Thought
Registration
Past Event
About Event
We'll be reading and discussing: Reasoning Models Struggle to Control their Chains of Thought
The paper asks a simple but important safety question: can reasoning models deliberately hide or reshape what they say in their chain-of-thought, and the authors find that they currently struggle to do that much more than they struggle to control their final answer.
Session Structure:
18:00-18:10: Introductions - please arrive on time!
18:10-18:50: silent paper reading
18:50-19:30: group discussion
This is a private event. If there is someone who you think would be a good fit for our community, please share this link with them.