5 papers
Matching Features, Not Tokens: Energy-Based Fine-Tuning of Language Models
Samy Jelassi, Mujin Kwun, Rosie Zhao +5
Cross-entropy (CE) training provides dense and scalable supervision for language models, but it optimizes next-token prediction under teacher forcing rather than sequence-level beh…
Random Scaling of Emergent Capabilities
Rosie Zhao, Tian Qin, David Alvarez-Melis +2
Language models famously improve under a smooth scaling law, but some specific capabilities exhibit sudden breakthroughs in performance. Advocates of "emergence" view these capabil…
Echo Chamber: RL Post-training Amplifies Behaviors Learned in Pretraining
Rosie Zhao, Alexandru Meterez, Sham Kakade +3
Reinforcement learning (RL)-based fine-tuning has become a crucial step in post-training language models for advanced mathematical reasoning and coding. Following the success of fr…
Deconstructing What Makes a Good Optimizer for Language Models
Rosie Zhao, Depen Morwani, David Brandfonbrener +2
Training language models becomes increasingly expensive with scale, prompting numerous attempts to improve optimization efficiency. Despite these efforts, the Adam optimizer remain…
SOAP: Improving and Stabilizing Shampoo using Adam
Nikhil Vyas, Depen Morwani, Rosie Zhao +5
There is growing evidence of the effectiveness of Shampoo, a higher-order preconditioning method, over Adam in deep learning optimization tasks. However, Shampoo's drawbacks includ…