AI Paper Atlas

Research map / Multi-agent systems and games / Human-AI and goal-conditioned

Off-Belief Learning

Hengyuan Hu, Adam Lerer, Brandon Cui, Luis Pineda, Noam Brown, Jakob Foerster

ICML 2021 · Multi-agent systems and games › Human-AI and goal-conditioned · approach: Boundedly rational agent modeling · 2 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (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 changedWhereMeasuredEffect
Apply OBL with a uniform-random starting policy to learn a grounded policyHanabi agents · 2-player HanabiCross-Play score: 20.85 $\pm$ 0.03 vs 21.77 $\pm$ 0.68Improved
Iterate OBL one level, using level 1 to generate beliefs (ablation)OBL · 2-player HanabiCross-Play score: 20.85 $\pm$ 0.03 -> 23.24 $\pm$ 0.03Improved
Iterate OBL from level 2 to level 3 (ablation)OBL · 2-player HanabiCross-Play score: 23.24 $\pm$ 0.03 -> 23.68 $\pm$ 0.05Improved
Iterate OBL from level 3 to level 4 (ablation)OBL · 2-player HanabiCross-Play score: 23.68 $\pm$ 0.05 -> 23.76 $\pm$ 0.06Improved
Iterate OBL from level 3 to level 4 (ablation)OBL · 2-player HanabiScore with Clone Bot: 16.54 $\pm$ 0.28 -> 16.76 $\pm$ 0.16Improved
Iterate OBL from level 2 to level 3 (ablation)OBL · 3-player HanabiScore with Clone Bot: 14.09 $\pm$ 0.48 -> 13.96 $\pm$ 0.47Hurt
Iterate OBL from level 3 to level 4 (ablation)OBL · 3-player HanabiScore with Clone Bot: 13.96 $\pm$ 0.47 -> 13.88 $\pm$ 0.47Hurt
Use OBL rather than Other-PlayHanabi agents · 3-player HanabiCross-Play score: 17.36 $\pm$ 0.19 -> 23.02 $\pm$ 0.01Improved
Use OBL rather than Other-PlayHanabi agents · 3-player HanabiScore with Clone Bot: 12.18 $\pm$ 0.25 -> 14.09 $\pm$ 0.48Improved
Add an auxiliary playable/discardable/unknown prediction task (ablation)Hanabi self-play baselines · 2-player HanabiSelf-Play score: 24.27 $\pm$ 0.01 vs 24.26 $\pm$ 0.01No clear effect
Add color-shuffling data augmentation (ablation)Clone Bot · 2-player HanabiSelf-Play score: 19.93 $\pm$ 0.09 -> 21.01 $\pm$ 0.07Improved

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy