Arthur Conmy

10 papers · 2023–2025 · 4 venues · across top CS/AI venues

Authors are matched by name, so researchers who share a name ("Yang Liu", "Wei Wang") appear as a single profile and their paper counts, venues and achievements are combined. How identity is handled.

Achievements

Jump to papers ↓

+3 more ↓

🐣 Hot Topic Early Bird 🌉 Interdisciplinary Bridge · Bronze 🌍 Conference Polyglot · Bronze 🔬 Deep Specialist (10) 👑 Triple Crown

🧭 Keyword Pioneer ⚡ Prolific Year · Productive 💎 Century Club · Bronze

Venues

EMNLP (3) ICML (3) ICLR (2) NeurIPS (2)

Top co-authors

Neel Nanda (5) Vikrant Varma (2) Can Rager (2) János Kramár (2) Tom Lieberum (2) Lewis Smith (2) Callum Stuart McDougall (2) Rohin Shah (2) Senthooran Rajamanoharan (2) Curt Tigges (1)

Research topics

Artificial Intelligence (10) Core AI (10) Machine Learning (5) Deep Learning (4) Core Methods (4) Natural Language Processing (2) Architectures (2) Resources & Methods (2) Learning Types (1) Optimization & Theory (1) Techniques (1)

Keywords

mechanistic interpretability (4) sparse autoencoder (4) language model (3) attention head (3) activation patching (2) transformer model (2) circuit discovery (2) neural network analysis (1) feature disentanglement (1) neural network optimization (1) latent representation (1) feature decomposition (1) task vector (1) causal intervention (1) black-box model (1) interpretable feature (1) token representation (1) reverse engineering (1) latent vector (1) vector arithmetic (1)

Papers

SAEBench: A Comprehensive Benchmark for Sparse Autoencoders in Language Model Interpretability ICML 2025 Scaling Sparse Feature Circuits For Studying In-Context Learning ICML 2025 Attribution Patching Outperforms Automated Circuit Discovery EMNLP 2024 Copy Suppression: Comprehensively Understanding a Motif in Language Model Attention Heads EMNLP 2024 Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2 EMNLP 2024 Improving Sparse Decomposition of Language Model Activations with Gated Sparse Autoencoders NeurIPS 2024 Stealing part of a production language model ICML 2024 Successor Heads: Recurring, Interpretable Attention Heads In The Wild ICLR 2024 Interpretability in the Wild: a Circuit for Indirect Object Identification in GPT-2 Small ICLR 2023 Towards Automated Circuit Discovery for Mechanistic Interpretability NeurIPS 2023