research
∙
02/06/2023
A Toy Model of Universality: Reverse Engineering How Networks Learn Group Operations
Universality is a key hypothesis in mechanistic interpretability – that ...
share
research
∙
01/12/2023
Progress measures for grokking via mechanistic interpretability
Neural networks often exhibit emergent behavior, where qualitatively new...
research
∙
12/21/2022
Language models are better than humans at next-token prediction
Current language models are considered to have sub-human capabilities at...
research
∙
05/03/2022
Adversarial Training for High-Stakes Reliability
In the future, powerful AI systems may be deployed in high-stakes settin...
research
∙
11/12/2021
Human irrationality: both bad and good for reward inference
Assuming humans are (approximately) rational enables robots to infer rew...
research
∙
04/23/2021
Optimal Cost Design for Model Predictive Control
Many robotics domains use some form of nonconvex model predictive contro...
research
∙
11/11/2020
Accounting for Human Learning when Inferring Human Preferences
Inverse reinforcement learning (IRL) is a common technique for inferring...
research
∙
01/24/2019