ICML 2024 Workshop MI Submissions

ReLU MLPs Can Compute Numerical Integration: Mechanistic Interpretation of a Non-linear Activation
Chun Hei Yip, Rajashree Agrawal, Jason Gross
- Published: 24 Jun 2024, Last Modified: 31 Jul 2024
- ICML 2024 MI Workshop Poster
- Readers: Everyone
Localizing Auditory Concepts in CNNs
Pratyaksh Gautam, Makarand Tapaswi, Vinoo Alluri
- Published: 24 Jun 2024, Last Modified: 31 Jul 2024
- ICML 2024 MI Workshop Poster
- Readers: Everyone
Interpreting Attention Layer Outputs with Sparse Autoencoders
Connor Kissane, Robert Krzyzanowski, Joseph Isaac Bloom, Arthur Conmy, Neel Nanda
- Published: 24 Jun 2024, Last Modified: 31 Jul 2024
- ICML 2024 MI Workshop Spotlight
- Readers: Everyone
Penzai + Treescope: A Toolkit for Interpreting, Visualizing, and Editing Models As Data
Daniel D. Johnson
- Published: 24 Jun 2024, Last Modified: 31 Jul 2024
- ICML 2024 MI Workshop Spotlight
- Readers: Everyone
How Do Transformers Fill in the Blanks? A Case Study on Matrix Completion
Pulkit Gopalani, Ekdeep Singh Lubana, Wei Hu
- Published: 24 Jun 2024, Last Modified: 31 Jul 2024
- ICML 2024 MI Workshop Poster
- Readers: Everyone
Dissecting Query-Key Interaction in Vision Transformers
Xu Pan, Aaron Philip, Ziqian Xie, Odelia Schwartz
- Published: 24 Jun 2024, Last Modified: 31 Jul 2024
- ICML 2024 MI Workshop Spotlight
- Readers: Everyone
Compact Proofs of Model Performance via Mechanistic Interpretability
Jason Gross, Rajashree Agrawal, Thomas Kwa, Euan Ong, Chun Hei Yip, Alex Gibson, Soufiane Noubir, Lawrence Chan
- Published: 24 Jun 2024, Last Modified: 31 Jul 2024
- ICML 2024 MI Workshop Spotlight
- Readers: Everyone
Modularity in Biologically Inspired Representations Depends on Task Variable Range Independence
Will Dorrell, Kyle Hsu, Luke Hollingsworth, Jin Hwa Lee, Jiajun Wu, Chelsea Finn, Peter E. Latham, Timothy Edward John Behrens, James C. R. Whittington
- Published: 24 Jun 2024, Last Modified: 24 Jun 2024
- ICML 2024 MI Workshop Poster
- Readers: Everyone
Does Editing Provide Evidence for Localization?
Zihao Wang, Victor Veitch
- Published: 24 Jun 2024, Last Modified: 31 Jul 2024
- ICML 2024 MI Workshop Poster
- Readers: Everyone
How Truncating Weights Improves Reasoning in Language Models
Lei Chen, Joan Bruna, Alberto Bietti
- Published: 24 Jun 2024, Last Modified: 31 Jul 2024
- ICML 2024 MI Workshop Poster
- Readers: Everyone
Mechanistic Interpretability of Binary and Ternary Transformer Networks
Jason Li
- Published: 24 Jun 2024, Last Modified: 31 Jul 2024
- ICML 2024 MI Workshop Poster
- Readers: Everyone
Learning Syntax Without Planting Trees: Understanding When and Why Transformers Generalize Hierarchically
Kabir Ahuja, Vidhisha Balachandran, Madhur Panwar, Tianxing He, Noah A. Smith, Navin Goyal, Yulia Tsvetkov
- Published: 24 Jun 2024, Last Modified: 31 Jul 2024
- ICML 2024 MI Workshop Spotlight
- Readers: Everyone
Controlling Large Language Model Agents with Entropic Activation Steering
Nate Rahn, Pierluca D'Oro, Marc G Bellemare
- Published: 24 Jun 2024, Last Modified: 31 Jul 2024
- ICML 2024 MI Workshop Poster
- Readers: Everyone
How Do Transformers "Do" Physics? Investigating the Simple Harmonic Oscillator
Subhash Kantamneni, Ziming Liu, Max Tegmark
- Published: 24 Jun 2024, Last Modified: 24 Jun 2024
- ICML 2024 MI Workshop Poster
- Readers: Everyone
Crafting Large Language Models for Enhanced Interpretability
Chung-En Sun, Tuomas Oikarinen, Tsui-Wei Weng
- Published: 24 Jun 2024, Last Modified: 31 Jul 2024
- ICML 2024 MI Workshop Poster
- Readers: Everyone
Identifying Functionally Important Features with End-to-End Sparse Dictionary Learning
Dan Braun, Jordan Taylor, Nicholas Goldowsky-Dill, Lee Sharkey
- Published: 24 Jun 2024, Last Modified: 31 Jul 2024
- ICML 2024 MI Workshop Spotlight
- Readers: Everyone
Challenges in Mechanistically Interpreting Model Representations
Satvik Golechha, James Dao
- Published: 24 Jun 2024, Last Modified: 31 Jul 2024
- ICML 2024 MI Workshop Poster
- Readers: Everyone
Comgra: A Tool for Analyzing and Debugging Neural Networks
Florian Dietz, Sophie Fellenz, Dietrich Klakow, Marius Kloft
- Published: 24 Jun 2024, Last Modified: 31 Jul 2024
- ICML 2024 MI Workshop Spotlight
- Readers: Everyone