AIbase
Product LibraryTool Navigation

scaling-monosemanticity-llama

Public

Reproducing Scaling Monosemanticity: Extracting Interpretable Features from Claude 3 Sonnet using LLaMA. This project explores monosemantic neurons in large language models, implementing and extending methods to scale and analyze interpretability in LLaMA-based architectures.

Creat2024-11-15T20:03:05
Update2025-03-20T21:31:30
4
Stars
0
Stars Increase