Showing cs.LGShow all
2 papers · 1 filter
cs.LG2026
HarnessMDP: Learning to Control LLM Agent Harnesses with Offline Reinforcement Learning
Haiwen Yi, Xinyuan Song
The execution harness is a critical component of LLM agents, shaping how the agent interacts with its environment and organizes multi-step execution. Prior work has shown that impr…
cs.LG2026
ManifoldFlow: Spectrum-Learnable Stiefel Optimization via SPD Relaxation
Haiwen Yi, Xinyuan Song
Orthogonal and Stiefel constraints are widely used in neural network optimization to control weight geometry, improve conditioning, and stabilize signal propagation. Yet standard S…