Research map / Bandits and online learning / Online convex optimization and games
AutoCFR: Learning to Design Counterfactual Regret Minimization Algorithms
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 AutoCFR using four extensive-form games instead of eight, including four normal-form games (ablation) | AutoCFR · Kuhn Poker | exploitability: 1.1489e-4 -> 5.7798e-5 | Helped |
| train AutoCFR using four extensive-form games instead of eight, including four normal-form games (ablation) | AutoCFR · Goofspiel (4) | exploitability: 1.0431e-6 -> 2.7402e-6 | Hurt |
| train AutoCFR using four extensive-form games instead of eight, including four normal-form games (ablation) | AutoCFR · Leduc Poker | exploitability: 1.4592e-6 -> 2.8801e-6 | Hurt |
| train AutoCFR using four extensive-form games instead of eight, including four normal-form games (ablation) | AutoCFR · HUNL Subgame (3) | exploitability: 2.2016e-5 -> 4.3293e-5 | Hurt |
| train AutoCFR using four extensive-form games instead of eight, including four normal-form games (ablation) | AutoCFR · HUNL Subgame (4) | exploitability: 9.8044e-6 -> 4.3926e-5 | Hurt |
| bootstrap the search with existing CFR variants rather than learning from scratch (ablation) | AutoCFR · Kuhn Poker | exploitability: 1.4492e-4 -> 1.1489e-4 | Helped |
| bootstrap the search with existing CFR variants rather than learning from scratch (ablation) | AutoCFR · Goofspiel (4) | exploitability: 3.3271e-4 -> 1.0431e-6 | Helped |
| bootstrap the search with existing CFR variants rather than learning from scratch (ablation) | AutoCFR · Leduc Poker | exploitability: 1.1364e-2 -> 1.4592e-6 | Helped |
| bootstrap the search with existing CFR variants rather than learning from scratch (ablation) | AutoCFR · HUNL Subgame (3) | exploitability: 4.8060e-3 -> 2.2016e-5 | Helped |
| bootstrap the search with existing CFR variants rather than learning from scratch (ablation) | AutoCFR · HUNL Subgame (4) | exploitability: 7.5901e-3 -> 9.8044e-6 | Helped |
Closest papers by meaning
- Solving Imperfect-Information Games via Discounted Regret Minimization (AAAI 2019 · similarity 0.91)
- Dynamic Discounted Counterfactual Regret Minimization (ICLR 2024 · similarity 0.91)
- Faster Game Solving via Asymmetry of Step Sizes (AAAI 2026 · similarity 0.91)
- Deep (Predictive) Discounted Counterfactual Regret Minimization (AAAI 2026 · similarity 0.91)
- Accelerating Nash Equilibrium Convergence in Monte Carlo Settings Through Counterfactual Value Based Fictitious Play (NeurIPS 2024 · similarity 0.89)
- Preference-CFR: Beyond Nash Equilibrium for Better Game Strategies (ICML 2025 · 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.