nowfound

Alternatives

Products that do what HyperSAE – Sparse Autoencoders, reimagined in hyperbolic space does

High-Performance Hyperbolic Sparse Autoencoders for Mechanistic Interpretability - vishal-dehurdle/hypersae

  1. 1

    High-Performance Hyperbolic Sparse Autoencoders for Mechanistic Interpretability - vishal-dehurdle/hypersae

    24d ago · github.com

  2. 2AG
  3. 3CT
  4. 4TA
  5. 5AS
  6. 6DL
  7. 7SA
  8. 8HM
  9. 9VA
  10. 10
    Sonnet134

    A new library for constructing neural networks from DeepMind

    2017

  11. 11SO
  12. 12

    Open-source machine learning library by Google

    2018

  13. 13LA
  14. 14AE
  15. 15DL
  16. 16AH
  17. 17RE
  18. 18HH
  19. 19AC
  20. 20PB
  21. 21MT
  22. 22AR
  23. 23IT
  24. 24KM

Ranked by how close each launch is in meaning, then by votes. Refine with a description →