Interpretability

11 posts · 3 sources

LangChainYouTube2w

What Happens Inside an AI Right Before It Cheats

AppleBlog3w

Dynamically Scaled Activation Steering

AnthropicBlog7 May

Natural Language Autoencoders

AnthropicBlog2 Apr

Emotion concepts and their function in a large language model

AnthropicBlog29 Oct

Emergent introspective awareness in large language models

AnthropicBlog29 May

Open-sourcing circuit-tracing tools

AnthropicBlog13 Mar

Auditing language models for hidden objectives

AnthropicBlog25 Oct

Evaluating feature steering: A case study in mitigating social biases

AnthropicBlog13 Jun

The engineering challenges of scaling interpretability

AnthropicBlog23 Apr

Simple probes can catch sleeper agents

AnthropicBlog17 Jun

Softmax Linear Units