activity
20242026
collaborators

12 papers

cs.LG2026

Neural Scaling Universality: If Exponents Are Fixed, Time to Understand Coefficients

Yizhou Liu, Jeff Gore

Neural scaling laws describe how pre-training loss decays as power laws with training time, model size, and compute. This position paper argues that the exponents of these power la…

cs.LG2026

Inverse Depth Scaling From Most Layers Being Similar

Yizhou Liu, Sara Kangaslahti, Ziming Liu +1

Neural scaling laws relate loss to model size in large language models (LLMs), yet depth and width may contribute to performance differently, requiring more detailed studies. Here,…

cs.LG2026

Universal One-third Time Scaling in Learning Peaked Distributions

Yizhou Liu, Ziming Liu, Cengiz Pehlevan +1

Training large language models (LLMs) is computationally expensive, partly because the loss exhibits slow power-law convergence whose origin remains debatable. Through systematic a…

cs.LG2026

Superposition unifies power-law training dynamics

Zixin Jessie Chen, Hao Chen, Yizhou Liu +1

We investigate the role of feature superposition in the emergence of power-law training dynamics using a teacher-student framework. We first derive an analytic theory for training…

cs.LG2026

The Blessing of Dimensionality in LLM Fine-tuning: A Variance-Curvature Perspective

Qiyao Liang, Jinyeop Song, Yizhou Liu +4

Weight-perturbation evolution strategies (ES) can fine-tune billion-parameter language models with surprisingly small populations (e.g., ), contradicting classical…

cs.LG2025

Superposition Yields Robust Neural Scaling

Yizhou Liu, Ziming Liu, Jeff Gore

The success of today's large language models (LLMs) depends on the observation that larger models perform better. However, the origin of this neural scaling law, that loss decrease…