research
∙
08/29/2023
Measurement Tampering Detection Benchmark
When training powerful AI systems to perform complex tasks, it may be ch...
share
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
∙
11/01/2022
Interpretability in the Wild: a Circuit for Indirect Object Identification in GPT-2 small
Research in mechanistic interpretability seeks to explain behaviors of m...
research
∙
10/04/2022
Polysemanticity and Capacity in Neural Networks
Individual neurons in neural networks often represent a mixture of unrel...
research
∙
05/03/2022
Adversarial Training for High-Stakes Reliability
In the future, powerful AI systems may be deployed in high-stakes settin...
research
∙
10/19/2018