Nora Belrose

26 papers A* 5Journal 21
YearRankTypeTitle / Venue / Authors
2025 A* conf
Automatically Interpreting Millions of Features in Large Language Models.
ICML
Gonçalo Paulo, Alex Mallen, Caden Juang, Nora Belrose
2025 J jnl
Binary Sparse Coding for Interpretability.
CoRR
Lucia Quirke, Stepan Shabalin, Nora Belrose
2025 J jnl
Converting MLPs into Polynomials in Closed Form.
CoRR
Nora Belrose, Alice Rigg
2025 A* conf
Do Transformer Interpretability Methods Transfer to RNNs?
AAAI
Gonçalo Paulo, Thomas Marshall, Nora Belrose
2025 J jnl
Estimating the Probability of Sampling a Trained Neural Network at Random.
CoRR
Adam Scherlis, Nora Belrose
2025 J jnl
Evaluating SAE interpretability without explanations.
CoRR
Gonçalo Paulo, Nora Belrose
2025 J jnl
Examining Two Hop Reasoning Through Information Content Scaling.
CoRR
David Johnston, Nora Belrose
2025 J jnl
Mechanistic Anomaly Detection for "Quirky" Language Models.
CoRR
David Johnston, Arkajyoti Chakraborty, Nora Belrose
2025 J jnl
Partially Rewriting a Transformer in Natural Language.
CoRR
Gonçalo Paulo, Nora Belrose
2025 J jnl
Slowing Learning by Erasing Simple Features.
CoRR
Lucia Quirke, Nora Belrose
2025 J jnl
Sparse Autoencoders Trained on the Same Data Learn Different Features.
CoRR
Gonçalo Paulo, Nora Belrose
2025 J jnl
Transcoders Beat Sparse Autoencoders for Interpretability.
CoRR
Gonçalo Paulo, Stepan Shabalin, Nora Belrose
2024 J jnl
Automatically Interpreting Millions of Features in Large Language Models.
CoRR
Gonçalo Paulo, Alex Mallen, Caden Juang, Nora Belrose
2024 J jnl
Balancing Label Quantity and Quality for Scalable Elicitation.
CoRR
Alex Mallen, Nora Belrose
2024 J jnl
Does Transformer Interpretability Transfer to RNNs?
CoRR
Gonçalo Paulo, Thomas Marshall, Nora Belrose
2024 A* conf
Neural Networks Learn Statistics of Increasing Complexity.
ICML
Nora Belrose, Quintin Pope, Lucia Quirke, Alex Mallen, Xiaoli Z. Fern
2024 J jnl
Neural Networks Learn Statistics of Increasing Complexity.
CoRR
Nora Belrose, Quintin Pope, Lucia Quirke, Alex Mallen, Xiaoli Z. Fern
2024 J jnl
Refusal in LLMs is an Affine Function.
CoRR
Thomas Marshall, Adam Scherlis, Nora Belrose
2024 J jnl
Understanding Gradient Descent through the Training Jacobian.
CoRR
Nora Belrose, Adam Scherlis
2023 A* conf
Adversarial Policies Beat Superhuman Go AIs.
ICML
Tony Tong Wang, Adam Gleave, Tom Tseng, Kellin Pelrine, Nora Belrose, Joseph Miller, Michael D. Dennis, Yawen Duan, Viktor Pogrebniak, Sergey Levine, Stuart Russell
2023 J jnl
Eliciting Latent Knowledge from Quirky Language Models.
CoRR
Alex Mallen, Nora Belrose
2023 J jnl
Eliciting Latent Predictions from Transformers with the Tuned Lens.
CoRR
Nora Belrose, Zach Furman, Logan Smith, Danny Halawi, Igor Ostrovsky, Lev McKinney, Stella Biderman, Jacob Steinhardt
2023 A* conf
LEACE: Perfect linear concept erasure in closed form.
NeurIPS
Nora Belrose, David Schneider-Joseph, Shauli Ravfogel, Ryan Cotterell, Edward Raff, Stella Biderman
2023 J jnl
LEACE: Perfect linear concept erasure in closed form.
CoRR
Nora Belrose, David Schneider-Joseph, Shauli Ravfogel, Ryan Cotterell, Edward Raff, Stella Biderman
2022 J jnl
Adversarial Policies Beat Professional-Level Go AIs.
CoRR
Tony Tong Wang, Adam Gleave, Nora Belrose, Tom Tseng, Joseph Miller, Michael D. Dennis, Yawen Duan, Viktor Pogrebniak, Sergey Levine, Stuart Russell
2022 J jnl
imitation: Clean Imitation Learning Implementations.
CoRR
Adam Gleave, Mohammad Taufeeque, Juan Rocamonde, Erik Jenner, Steven H. Wang, Sam Toyer, Maximilian Ernestus, Nora Belrose, Scott Emmons, Stuart Russell