Naomi Saphra

PolyPythias: Stability and Outliers across Fifty Language Model Pre-Training Runs

arXiv 2025

Sometimes I am a Tree: Data Drives Unstable Hierarchical Generalization

arXiv 2024

Recite, Reconstruct, Recollect: Memorization in LMs as a Multifaceted Phenomenon

arXiv 2024

Sudden Drops in the Loss: Syntax Acquisition, Phase Transitions, and Simplicity Bias in MLMs

arXiv 2023

Latent State Models of Training Dynamics

arXiv 2023

TRAM: Bridging Trust Regions and Sharpness Aware Minimization

arXiv 2023

The MultiBERTs: BERT Reproductions for Robustness Analysis

the-multiberts-bert-reproductions-for-1

Angelica Chen

Kyunghyun Cho

professor

Stella Biderman

founder

Alexander D'Amour

Alvin Deng

Christopher A. Choquette-Choo

David Alvarez-Melis

Dipanjan Das

Ellie Pavlick

Hailey Schoelkopf