Showing cs.LGShow all
3 papers · 1 filter
cs.LG2026
An Isotropy-Preserving Spectral Cap for Muon: Theory and Three Case Studies
Jiachun Li
Muon and related matrix-sign optimizers are increasingly used to pre-train large language models, but their effect on the internal geometry of individual weight matrices is not wel…
cs.LG2026
Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation
Jiachun Li, David Simchi-Levi
Adaptive experiments for average treatment effects (ATE) require randomized allocations balancing valid inference with statistical efficiency. The oracle design is a covariate-depe…
cs.LG2025
Beyond Covariance Matrix: The Statistical Complexity of Private Linear Regression
Fan Chen, Jiachun Li, Alexander Rakhlin +1
We study the statistical complexity of private linear regression under an unknown, potentially ill-conditioned covariate distribution. Somewhat surprisingly, under privacy constrai…