Research map / Bandits and online learning / Online convex optimization and games
Efficient Deviation Types and Learning for Hindsight Rationality in Extensive-Form 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 |
|---|---|---|---|
| Replace informed action deviations with full behavioral deviations (ablation) | EFR · Goofspiel (2-player), fixed | win frequency: 0.51 -> 0.63 | Helped |
| Replace informed action deviations with full behavioral deviations (ablation) | EFR · Sheriff, fixed | points: 0.28 -> 0.91 | Helped |
| Use TIPS instead of informed action deviations (ablation) | EFR · Goofspiel (2-player), fixed | win frequency: 0.51 -> 0.60 | Helped |
| Use TIPS instead of informed action deviations (ablation) | EFR · Sheriff, fixed | points: 0.28 -> 0.82 | Helped |
| Use full behavioral deviations instead of TIPS (ablation) | EFR · Goofspiel (2-player), fixed | win frequency: 0.60 -> 0.63 | Helped |
| Use full behavioral deviations instead of TIPS (ablation) | EFR · Sheriff, fixed | points: 0.82 -> 0.91 | Helped |
| Use full behavioral deviations instead of TIPS (ablation) | EFR · Goofspiel (3-player), fixed | win frequency: 0.53 -> 0.53 | No clear effect |
| Use full behavioral deviations instead of TIPS (ablation) | EFR · Sheriff, simultaneous | points: 0.38 -> 0.38 | No clear effect |
| Use TIPS instead of CSPS (ablation) | EFR · Goofspiel (3-player), simultaneous | win frequency: 0.91 -> 0.87 | Hurt |
| Use TIPS instead of CSPS (ablation) | EFR · Sheriff, fixed | points: 0.61 -> 0.82 | Helped |
| Use stronger deviation types rather than weaker types (ablation) | EFR variants · Benchmark games | expected payoff | Mixed |
Closest papers by meaning
- Hindsight and Sequential Rationality of Correlated Play (AAAI 2021 · similarity 0.93)
- Optimism Without Regularization: Constant Regret in Zero-Sum Games (NeurIPS 2025 · similarity 0.87)
- Learning from Delayed Feedback in Games via Extra Prediction (NeurIPS 2025 · similarity 0.87)
- Mediator Interpretation and Faster Learning Algorithms for Linear Correlated Equilibria in General Sequential Games (ICLR 2024 · similarity 0.87)
- Convergence of No-Swap-Regret Dynamics in Self-Play (NeurIPS 2024 · similarity 0.87)
- The impact of uncertainty on regularized learning in games (ICML 2025 · similarity 0.87)
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.