← Latest products

At 16K features

flat autoencoders break. Curved space doesn't. High-Performance Hyperbolic Sparse Autoencoders for Mechanistic Interpretability.

Open source

13 stars · GitHubexternal site

Visit github.comexternal site

Reactions (1)

Similar products

by description similarity

  1. Llama 3.3 70B Sparse Autoencoders with API access

    goodfire.ai

    Hacker News201 pts

  2. Autodistill

    github.com

    Use big slow foundation models to train small fast models

    Hacker News18 pts

  3. Llama 3.2 Interpretability with Sparse Autoencoders

    github.com

    A complete end-to-end pipeline for LLM interpretability with sparse autoencoders (SAEs) using Llama 3.2, written in pure PyTorch and fully reproducible.

    Hacker News579 pts

  4. TinyGPT

    github.com

    A Simple, Educational Deep Learning Library

    Hacker News13 pts

  5. 40M embeddings to find who knows what on HN

    blog.wilsonl.in

    Hacker News33 pts