RookWorld-LM Chain-of-Thought Chess

Strategic reasoning with transparent thought processes • 32.1% checkmate-in-one • Outperforms ChessGPT-Base 26.5% with 24x fewer parameters • 124M unified agent+environment

Jonathan Rahn • Jenia Jitsev • Qi Sun • LAION Research Project

Loading RookWorld-LM

Initializing...

Loading model from HuggingFace (~285MB). First load may take 60-90 seconds...

Chain-of-Thought Reasoning

Watch RookWorld-LM analyze positions step-by-step with transparent reasoning traces

Move: 1
Turn: White
Status: Ready to start
rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq - 0 1

Live Reasoning

Ready to start self-play

Click "Start Self-Play" to watch RookWorld-LM play chess with transparent reasoning.

You'll see:

  • P: Position analysis streaming in real-time
  • M: Candidate moves being considered
  • E: Evaluation of each candidate
  • B: Best move selection
  • Environment validation: State transition verification

Candidate Moves

Start self-play to see candidate moves...