Neel Nanda's Classic Paper Decomposes Grokking into Three Stages

Neel Nanda's classic 2023 paper with Lawrence Chan and Jacob Steinhardt is being reshared, using mechanistic interpretability to decompose grokking into three progressive training stages.

2026-11-10 ~ 2026-11-10 · 2 related posts

• Neel Nanda's grokking paper: reverse-engineering emergent behavior into three phases — burny_tech · 2026-11-10
• Neel Nanda's Classic Paper: Mechanistically Explaining Grokking in Three Phases — burny_tech · 2026-11-10