Saida Liu

mail · GitHub

I am an undergraduate at Kobe University, currently an exchange student at TUM (2025/26). My interests lie in safe and general physical intelligence, especially Diffusion Models and Safe Reinforcement Learning (Safe RL).

News

Selected Projects & Preprints

• MATT-Diff: Multimodal Active Target Tracking by Diffusion Policy (paper, code) — Submitted to L4DC, preprint on arXiv.

Open‑source

• RL note 1 RL from an optimization perspective
• FSNet‑test — a simple SafeRL like differentiable safe set projection demos.