Shaocong Ma

New Hybrid Fine-Tuning Paradigm for LLMs: Algorithm Design and Convergence Analysis Framework

Riemannian Zeroth-Order Gradient Estimation with Structure-Preserving Metrics for Geodesically Incomplete Manifolds

A Multi-Fidelity Mixture-of-Expert Framework Integrating PDE Solvers and Neural Operators for Computational Fluid Dynamics

On the Optimal Construction of Unbiased Gradient Estimators for Zeroth-Order Optimization

Revisiting Zeroth-Order Optimization: Minimum-Variance Two-Point Estimators and Directionally Aligned Perturbations

Robust Reinforcement Learning in Finance: Modeling Market Impact with Elliptic Uncertainty Sets

Hybrid Fine-Tuning of LLMs: Theoretical Insights on Generalized Smoothness and Convergence

Rectified Robust Policy Optimization for Robust Constrained Reinforcement Learning without Strong Duality