Training Is Nothing Like Learning with Naomi Saphra (Harvard)

The Information Bottleneck · Mar 13, 2026 · 1h 11m · transcript

Naomi Saphra, Kempner Research Fellow at Harvard and incoming Assistant Professor at Boston University, joins us to explain why you can't do interpretability without understanding training dynamics, in the same way you can't do biology without evolution.

Naomi argues that many structures researchers find inside trained models are vestigial, they mattered early in training but are meaningless by the end. Grokking is one case of a broader phenomenon: models go through multiple consecutive phase transitions during training, driven by symmetry breaking and head specialization, but the smooth loss…