Adam Karvonen

Expert
@adamkarvonen
adamkarvonen.github.io/about/

chess_llm_interpretability. Visualizing the internal board state of a GPT trained on chess PGN strings, and performing interventions on its internal board state and representation of player Elo.

220

SAEBench. Python

179

activation_oracles. Python

95

chess_gpt_eval. A repo to evaluate various LLM's chess playing abilities.

91

train_ChessGPT. A repository for training nanogpt-based Chess playing language models.

30

dictionary_learning_demo. Jupyter Notebook

26

SAE_BoardGameEval. Python

25

difr. Python

6

llm_bias. Python

5

sae_kl_finetune. Python

4

interp_tools. Python

3

OthelloUnderstanding. Jupyter Notebook

2

token-difr. Python

2

rl_oocr. Python

1

dictionary_learning. Python

1

trl_train_demo. Python

1

whitebox_evals. Jupyter Notebook

1
17
Apply