Research map / Multi-agent systems and games / Games and equilibria
Guarantees for Self-Play in Multiplayer Games via Polymatrix Decomposability
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 |
|---|---|---|---|
| Approximate the strategy neighborhood with a subgame-stable constant-sum polymatrix game | CFR+-learned strategies in 3-player Leduc poker · 3-player Leduc poker | δ (CSP approximation error): δ=0.009±0.00046 | Helped |
| Approximate the strategy neighborhood with a subgame-stable constant-sum polymatrix game | CFR+-learned strategies in 3-player Leduc poker · 3-player Leduc poker | γ (subgame-stability error): γ=0.004±0.00016 | Helped |
| Use the δ and γ decomposition-derived bound to predict vulnerability | CFR+-learned strategies in 3-player Leduc poker · 3-player Leduc poker | vulnerability bound relative to measured vulnerability: 1.89 times ... 3.05 times ... 2.51 times ... standard error of 0.049 | Mixed |
| Use CFR+ instead of the previously reported CFR result | 3-player Leduc poker · 3-player Leduc poker | ε-Nash exploitability: 0.130 -> ε≤0.013 | Helped |
| Apply CFR to 3-player Leduc poker | CFR · 3-player Leduc poker | ε-Nash exploitability: ε ≤ 0.017 | Helped |
| Evaluate a toy Hanabi game with SGDecompose | Self-play-learned strategies · Toy Hanabi | δ and γ | Hurt |
Closest papers by meaning
- Securing Equal Share: A Principled Approach for Learning Multiplayer Symmetric Games (ICML 2025 · similarity 0.90)
- The Complexity of Two-Team Polymatrix Games with Independent Adversaries (ICLR 2025 · similarity 0.89)
- Convergence of No-Swap-Regret Dynamics in Self-Play (NeurIPS 2024 · similarity 0.88)
- Fast Last-Iterate Convergence of Learning in Games Requires Forgetful Algorithms (NeurIPS 2024 · similarity 0.88)
- From Average-Iterate to Last-Iterate Convergence in Games: A Reduction and Its Applications (NeurIPS 2025 · similarity 0.88)
- Efficient Learning in Polyhedral Games via Best-Response Oracles (AAAI 2024 · similarity 0.88)
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.