| | Natural Language Autoencoders Produce Unsupervised Explanations LLM Activation (transformer-circuits.pub) |
| 4 points by Anon84 12 days ago | past | discuss |
|
| | NL Autoencoders Produce Unsupervised Explanations of LLM Activations (transformer-circuits.pub) |
| 3 points by rajeevn 16 days ago | past |
|
| | HeadVis: An Interactive Tool for Investigating Attention Heads (transformer-circuits.pub) |
| 3 points by MrOrelliOReilly 18 days ago | past |
|
| | HeadVis: An Interactive Tool for Investigating Attention Heads (transformer-circuits.pub) |
| 4 points by rajeevn 19 days ago | past |
|
| | Emotion Concepts and Their Function in a Large Language Model (transformer-circuits.pub) |
| 63 points by Anon84 49 days ago | past | 10 comments |
|
| | Emotion Concepts and Their Function in a Large Language Model (transformer-circuits.pub) |
| 6 points by majkinetor 50 days ago | past |
|
| | Emotion Concepts and Their Function in a Large Language Model (transformer-circuits.pub) |
| 3 points by stared 51 days ago | past |
|
| | Anthropic's Interpretability Research Blog (transformer-circuits.pub) |
| 3 points by philipfweiss 4 months ago | past | 1 comment |
|
| | Emergent introspective awareness in large language models (transformer-circuits.pub) |
| 1 point by lawrenceyan 6 months ago | past |
|
| | Emergent Introspective Awareness in Large Language Models (transformer-circuits.pub) |
| 30 points by famouswaffles 6 months ago | past | 4 comments |
|
| | When models manipulate manifolds: The geometry of a counting task (transformer-circuits.pub) |
| 98 points by vinhnx 6 months ago | past | 17 comments |
|
| | When Models Manipulate Manifolds: The Geometry of a Counting Task (transformer-circuits.pub) |
| 4 points by 1wheel 6 months ago | past |
|
| | Visual Features Across Modalities: SVG and ASCII Art Cross-Modal Understanding (transformer-circuits.pub) |
| 12 points by vismit2000 7 months ago | past | 1 comment |
|
| | LLMs extract high-level semantic concepts from SVG and ASCII art (transformer-circuits.pub) |
| 3 points by neuronerd1 7 months ago | past | 1 comment |
|
| | When Models Manipulate Manifolds: The Geometry of a Counting Task (transformer-circuits.pub) |
| 2 points by tanelpoder 7 months ago | past |
|
| | When Models Manipulate Manifolds: The Geometry of a Counting Task (transformer-circuits.pub) |
| 5 points by e_ameisen 7 months ago | past |
|
| | Transformer Circuits: reverse-engineering transformers into graspable programs (transformer-circuits.pub) |
| 1 point by dvrp 10 months ago | past |
|
| | So You Want to Work in Mechanistic Interpretability? (transformer-circuits.pub) |
| 2 points by jxmorris12 11 months ago | past |
|
| | Circuit Tracing: Revealing Computational Graphs in Language Models (Anthropic) (transformer-circuits.pub) |
| 173 points by ydnyshhh on March 31, 2025 | past | 27 comments |
|
| | The Biology of a Large Language Model (transformer-circuits.pub) |
| 117 points by frozenseven on March 28, 2025 | past | 19 comments |
|
| | Circuit Tracing: Revealing Computational Graphs in Language Models (transformer-circuits.pub) |
| 8 points by mfiguiere on March 27, 2025 | past |
|
| | The Biology of a Large Language Model (transformer-circuits.pub) |
| 3 points by mfiguiere on March 27, 2025 | past |
|
| | Insights on Cross-Coder Model Diffing (transformer-circuits.pub) |
| 1 point by gregorymichael on Feb 24, 2025 | past |
|
| | Transformer Circuits Thread (transformer-circuits.pub) |
| 1 point by fzliu on Feb 5, 2025 | past |
|
| | Definitions and Motivation: Features, Directions, and Superposition (transformer-circuits.pub) |
| 4 points by Bluestein on Dec 27, 2024 | past |
|
| | Toy Models of Superposition (2022) (transformer-circuits.pub) |
| 45 points by tessierashpool9 on Nov 8, 2024 | past |
|
| | Transformer Circuits Thread (transformer-circuits.pub) |
| 2 points by plurby on Nov 3, 2024 | past |
|
| | Sparse Crosscoders for Cross-Layer Features and Model Diffing (transformer-circuits.pub) |
| 2 points by benocodes on Oct 25, 2024 | past |
|
| | A collection of small updates from the Anthropic Interpretability team (transformer-circuits.pub) |
| 2 points by daralthus on July 31, 2024 | past |
|
| | Scaling Monosemanticity: Extracting Interpretable Features from Claude 3 Sonnet (transformer-circuits.pub) |
| 22 points by Anon84 on May 23, 2024 | past | 1 comment |
|
|
| More |
