apartresearch/deepdecipher
🦠 DeepDecipher: An open source API to MLP neurons
DeepDecipher is a tool for researchers and practitioners interested in understanding how Large Language Models (LLMs) work internally. It helps you examine individual "neurons" within these models to see what concepts or features they respond to. You provide the model and neuron you're curious about, and it delivers detailed explanations, activation patterns, and visual insights into that neuron's function. This is designed for anyone doing mechanistic interpretability of LLMs.
No commits in the last 6 months.
Use this if you need to deeply inspect and understand the specific internal workings of a large language model, particularly how individual neurons contribute to its overall behavior.
Not ideal if you're looking for a tool to simply fine-tune, deploy, or interact with LLMs at a high level without diving into their internal structure.
Stars
9
Forks
—
Language
Rust
License
MIT
Last pushed
May 02, 2024
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/apartresearch/deepdecipher"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
obss/sahi
Framework agnostic sliced/tiled inference + interactive ui + error analysis plots
tensorflow/tcav
Code for the TCAV ML interpretability project
MAIF/shapash
🔅 Shapash: User-friendly Explainability and Interpretability to Develop Reliable and Transparent...
TeamHG-Memex/eli5
A library for debugging/inspecting machine learning classifiers and explaining their predictions
csinva/imodels
Interpretable ML package 🔍 for concise, transparent, and accurate predictive modeling...