AI Paper Atlas

Research map / Multi-agent systems and games / Cooperative multi-agent RL

Scaling Multi-Agent Reinforcement Learning with Selective Parameter Sharing

Filippos Christianos, Georgios Papoudakis, Muhammad A Rahman, Stefano V Albrecht

ICML 2021 · Multi-agent systems and games › Cooperative multi-agent RL · approach: MARL benchmarks and grouping · 17 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
Selectively share parameters among clustered agents instead of giving each agent an independent policyA2C MARL · BPSsum of returns: -189.99 -> -179.69; -385.18 -> -369.69; -401.09 -> -371.99; -388.21 -> -378.50Improved
Selectively share parameters among clustered agents instead of giving each agent an independent policyA2C MARL · BPS-hsum of returns: -187.92 -> -189.93; -398.11 -> -373.92Mixed
Selectively share parameters among clustered agents instead of using SAPIA2C MARL · BPS-h (3)sum of returns: -4085.51 -> -2522.61Improved
Selectively share parameters among clustered agents instead of using independent policiesA2C MARL · C-RWAREsum of returns: 2.93 -> 7.03; 0.28 -> 20.88Improved
Selectively share parameters among clustered agents instead of using SAPIA2C MARL · C-RWARE (3)sum of returns: 21.30 -> 32.27Improved
Selectively share parameters among clustered agents instead of independent policiesA2C MARL · LBFsum of returns: 0.91 -> 0.97Improved
Share parameters selectively rather than using SAPIA2C MARL · MMM2sum of returns: 168.39 -> 179.45Improved
Use fully shared parameters instead of independent policiesA2C MARL · MMM2sum of returns: 72.32 -> 128.75Improved
Use fully shared parameters instead of independent policiesA2C MARL · BPS and C-RWAREsum of returns: -189.99 -> -612.18; 2.93 -> 0.57Hurt
Use independent policies instead of parameter sharingA2C MARL · C-RWARE (2) and (3)sum of returns: 20.88 -> 0.28; 32.27 -> 0.33Hurt
Scale up SAPI network width with the number of BPS-h colours (ablation)SAPI · BPS-hlearning successNo clear effect
Form fewer clusters than needed (ablation)OPS · C-RWAREsum of returnsHurt

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