Changxu Wei — accepted AI-conference papers

2026

Policy Newton Algorithm in Reproducing Kernel Hilbert Space

Yixian Zhang, Huaze Tang, Changxu Wei, Chao Wang, Wenbo Ding

ICLR 2026poster

Reinforcement learning (RL) policies represented in Reproducing Kernel Hilbert Spaces (RKHS) offer powerful representational capabilities. While second-order optimization methods like Newton's method demonstrate faster convergence than first-order approaches, current RKHS-based policy optimization r…

Cited by 0 Source ↗Scholar ↗

2025

Distributional Decision Transformer: Risk-Sensitive Offline RL via Quantile-Based Critics and Stochastic Return

Changxu Wei, Huaze Tang, Yixian Zhang, Chao Wang, Xiao-Ping Zhang, Wenbo Ding

IROS 2025

Offline reinforcement learning faces a critical challenge in synthesizing high-reward trajectories from suboptimal datasets while robustly handling the stochasticity inherent in real-world decision-making. While combination of return-conditioned sequence models, such as Decision Transformers (DT), a

Cited by 0 Source ↗Scholar ↗