Models and datasets used in the paper "Interpreting Reasoning Features in Large Language Models via Sparse Autoenoder": https://arxiv.org/abs/2503.188
-
andreuka18/DeepSeek-R1-Distill-Llama-8B-lmsys-openthoughts-tokenized
Viewer • Updated • 781k • 150 -
andreuka18/deepseek-r1-distill-llama-8b-lmsys-openthoughts
Text Generation • Updated -
andreuka18/OpenThoughts-10k-DeepSeek-R1
Viewer • Updated • 10k • 13 -
I Have Covered All the Bases Here: Interpreting Reasoning Features in Large Language Models via Sparse Autoencoders
Paper • 2503.18878 • Published • 119
