AI Paper Atlas

Research map / Multi-agent systems and games / Games and equilibria

Joint Policy Search for Multi-agent Collaboration with Imperfect Information

Yuandong Tian, Qucheng Gong, Yu Jiang

NeurIPS 2020 · Multi-agent systems and games › Games and equilibria · approach: Learning in Markov games · 7 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 tabular JPS to CFR1kCFR1k policies · Simple collaborative gamesaverage reward: CFR1k 0.85 -> CFR1k+JPS 1.00 (Comm L=5); CFR1k 2.60 -> CFR1k+JPS 2.74 (2-Suit Mini-Bridge N=5)Improved
Jointly search the current player's and partner's policiesA2C baseline · Contract Bridge bidding vs 10 independent A2C baselinesIMPs/b: 0.27 $\pm$ 0.13 -> 0.66 $\pm$ 0.11Improved
Increase the fraction of games using JPS from 5% to 10% (ablation)JPSBid · Contract Bridge bidding vs 10 independent A2C baselinesIMPs/b: 0.66 $\pm$ 0.11 -> 0.39 $\pm$ 0.11Hurt
Use single-player search instead of joint policy search (ablation)A2C baseline · Contract Bridge bidding vs baselineIMPs/b: 0.20 -> 0.46Improved
Add explicit belief-modeling auxiliary loss (ablation)A2C baseline · Contract Bridge biddingIMPs: 2.99 $\pm$ 0.04 -> 2.53 $\pm$ 0.27Hurt
Increase network blocks from 2 to 4 (ablation)A2C baseline · Contract Bridge biddingIMPs: 2.97 $\pm$ 0.05 -> 2.99 $\pm$ 0.04No clear effect
Increase sync frequency from 1 to 50 (ablation)A2C baseline · Contract Bridge biddingIMPs: 2.89 $\pm$ 0.13 -> 2.99 $\pm$ 0.04No clear effect
Add random simultaneous exploration by the agent and partner (ablation)A2C baseline · Contract Bridge biddingIMPs: 2.99 $\pm$ 0.04 -> 2.37 $\pm$ 0.31Hurt
Use 1 sampled state per infoset rather than all states (ablation)CFR1k+JPS · Mini-Hanabiaverage reward: 9.50 -> 10.00Improved
Increase sampled states per infoset from 1 to 15 (ablation)CFR1k+JPS · Mini-Hanabiaverage reward: 10.00 -> 9.51Hurt
Use sample-based JPS rather than all-state JPS (ablation)CFR1k initialization · Simple Bidding N=16average reward: 10.56 -> 10.61Improved
Use sample-based JPS rather than all-state JPS (ablation)BAD initialization · Simple Bidding N=16average reward: 10.47 -> 9.91 (1 sample); 10.47 -> 10.57 (25 samples)Mixed

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