videlalvaro/llm-arithmetic-internals
Mechanistic experiments on how LLMs represent and compute arithmetic internally, with strict no-parser controls, reproducible audits, and an interactive article.
GitHub repository with 6 stars and 0 forks.
Language: Python
Topics: arithmetic, interpretability, llm, mechanistic-interpretability, probes, residual-stream, sae, sparse-autoencoder, transformers