Pinned Loading

1. AlphaXiang AlphaXiang Public

   AlphaZero-style XiangQi(Chinese Chess) engine based on MCTS + PyTorch (support Apple Silicon MPS). Reproducible self-playing, checkpoint auto-resume, and a PVP client with clocks/undo. Built like a…

   Python 2

2. AlphaDiablo AlphaDiablo Public

   Fast, deterministic Diablo I RL environment on DevilutionX + a PPO pipeline with a fully documented ten-run lesson chain

   Python 1

3. aispeedrun aispeedrun Public

   HTML

4. AlphaZuma AlphaZuma Public

   Fidelity-first Zuma's Revenge RL environment and AlphaZuma-55 learned speedrun agents: tick-exact simulator, preregistered experiments with receipts

   Python

5. kingdom-rush-ml-agent kingdom-rush-ml-agent Public

   An ML agent that cleared the full 12-level main campaign of the original Kingdom Rush (Normal, new save): 5/5 on held-out final seeds, with verifiable evidence

   Python