activity
20242026
collaborators

18 papers

cs.LG2026

Bergson: An Open Source Library for Data Attribution

Lucia Quirke, Louis Jaburi, David Johnston +6

Data attribution is a promising field in interpretability that aims to explain model behavior through the influence of its training data, with applications including debugging unde…

cs.LG2025

Eliciting Latent Predictions from Transformers with the Tuned Lens

Nora Belrose, Igor Ostrovsky, Lev McKinney +5

We analyze transformers from the perspective of iterative inference, seeking to understand how model predictions are refined layer by layer. To do so, we train an affine probe for…

cs.LG2025

Binary Sparse Coding for Interpretability

Lucia Quirke, Stepan Shabalin, Nora Belrose

Sparse autoencoders (SAEs) are used to decompose neural network activations into sparsely activating features, but many SAE features are only interpretable at high activation stren…

cs.LG2025

Automatically Interpreting Millions of Features in Large Language Models

Gonçalo Paulo, Alex Mallen, Caden Juang +1

While the activations of neurons in deep neural networks usually do not have a simple human-understandable interpretation, sparse autoencoders (SAEs) can be used to transform these…

cs.LG2025

Evaluating SAE interpretability without explanations

Gonçalo Paulo, Nora Belrose

Sparse autoencoders (SAEs) and transcoders have become important tools for machine learning interpretability. However, measuring how interpretable they are remains challenging, wit…

cs.LG2025

Mechanistic Anomaly Detection for "Quirky" Language Models

David O. Johnston, Arkajyoti Chakraborty, Nora Belrose

As LLMs grow in capability, the task of supervising LLMs becomes more challenging. Supervision failures can occur if LLMs are sensitive to factors that supervisors are unaware of.…