CJC Chess RL — Trained Agent vs Random
Actor-Critic [288→128→128→P+V] | A2C + GAE | 300 Episodes Trained | Zero Dependencies
Move 0 / 80
White (Agent) to move
Agent: White
Opponent: Random
Seed: 42424
Training: 300 episodes (200 vs Random + 100 vs Heuristic)
Network: ~110K params | Features: 288-dim | Manual backprop