Stella Biderman 16/03/2023 Stella Biderman 16/03/2023

Interpreting Across Time

How do properties of models emerge and evolve over the course of training?

Stella Biderman 15/03/2023 Stella Biderman 15/03/2023

Eliciting Latent Knowledge

As models get smarter, humans won't always be able to independently check if a model's claims are true or false. We aim to circumvent this issue by directly eliciting latent knowledge (ELK) inside the model’s activations.

Stella Biderman 16/02/2023 Stella Biderman 16/02/2023

Training LLMs

EleutherAI has trained and released many powerful open source LLMs.

Stella Biderman 15/02/2023 Stella Biderman 15/02/2023

Evaluating LLMs

Evaluating advanced AI models in robust and reliable ways.

Stella Biderman 14/02/2023 Stella Biderman 14/02/2023

Alignment MineTest

Alignment-MineTest is a research project that uses the open source Minetest voxel engine as a platform for studying AI alignment.

Stella Biderman 13/02/2023 Stella Biderman 13/02/2023

Mesaoptimization

Studying how auxiliary optimization objectives arise in models

Stella Biderman 22/12/2022 Stella Biderman 22/12/2022

Polyglot

Building LLMs and doing NLP in non-English languages.