Research map / Multi-agent systems and games / Games and equilibria
Combining Deep Reinforcement Learning and Search for Imperfect-Information Games
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| train value network on random public belief states instead of self-play data (ablation) | ReBeL · Turn endgame hold'em | exploitability | Hurt |
| use ReBeL (CFR-D) instead of full-game FP | Liar's Dice 2x3f · Liar's Dice 2x3f | exploitability: 0.057 -> 0.017 | Helped |
| use full-game tabular CFR instead of ReBeL | ReBeL · Liar's Dice | exploitability: 0.017 -> 0.001 in favor of tabular CFR at 1x4f | Hurt |
| use ReBeL instead of Modicum | Modicum · HUNL vs Slumbot | win rate: 11 ± 5 -> 45 ± 5 | Helped |
| use ReBeL instead of DeepStack against LBR | DeepStack · HUNL vs LBR | win rate: 383 ± 112 -> 881 ± 94 | Helped |
| replace MSE with Huber loss for value network (ablation) | ReBeL · preliminary experiments | value accuracy | Helped |
Closest papers by meaning
- Reevaluating Policy Gradient Methods for Imperfect-Information Games (ICLR 2026 · similarity 0.90)
- Depth-Limited Solving for Imperfect-Information Games (NeurIPS 2018 · similarity 0.89)
- Stabilizing RL+Search for Imperfect-Information Extensive-Form Games (NeurIPS 2026 · similarity 0.88)
- Outbidding and Outbluffing Elite Humans: Mastering Liar’s Poker via Self-Play and Reinforcement Learning (NeurIPS 2026 · similarity 0.88)
- GAE Falls Short in Imperfect-Information Self-Play Reinforcement Learning (NeurIPS 2026 · similarity 0.87)
- Mastering Zero-Shot Interactions in Cooperative and Competitive Simultaneous Games (ICML 2024 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.