From ideas to experiments. From experiments to tools.

I'm HunLuanZhiZhu, a graduate student exploring the intersection of AI research and software engineering. My work spans optimization, reinforcement learning, spiking neural networks, agent systems and interactive browser experiments.

I prefer testable ideas, working prototypes and reusable tools over long lists of buzzwords.

Research visualizations, interactive graphics, games and utilities live at zyh.sryze.cc. Public site source: hunluanzhizhu.github.io.

Each illustrated card links to a public GitHub repository.

Adaptive fractional-gradient optimizers.	Parallel reset mechanisms for spiking neurons.
Model API proxy with streaming and request adaptation.	Reusable agent skills and research workflows.
Multi-agent game development plugin.	Game creation workflows adapted for ZCode.
Research presentations published for the browser.	Interactive calibration using video and graphics.

Only project names are shown below. No repository links, implementation details or project progress are disclosed.

MaxFormer  ·  Guon  ·  alo-marl-fs  ·  math-modeling-skills  ·  pnn  ·  mems  ·  MUD3  ·  rl-fs  ·  MinimalWorldModel