Atticus Geiger

Atticus Geiger on Hugging Face Daily Papers: 5 papers, 1 in the top 3 of their day, 138 upvotes.

1. Decomposing MLP Activations into Interpretable Features via Semi-Nonnegative Matrix Factorization 6 upvotes, #27 of 2025-06-13
2. Open Problems in Mechanistic Interpretability 16 upvotes, #5 of 2025-01-29
3. Enhancing Automated Interpretability with Output-Centric Feature Descriptions 10 upvotes, #13 of 2025-01-15
4. ReFT: Representation Finetuning for Language Models 58 upvotes, #1 of 2024-04-05
5. Interpretability at Scale: Identifying Causal Mechanisms in Alpaca 2 upvotes, #8 of 2023-05-16

Data: hysts-bot-data/daily-papers-stats and the Daily Papers API. Open data: tardellirs/paper-pulse-data. Sister project: Model Pulse, the download history of every model on the Hub.