AI Science Tracker
EN

From Single-Agent Self-Discovery to Multi-Agent Adaptation: Scaling Laws of Partner Feedback and Convention Rigidity in Meta Reinforcement Learning Algorithms

Not stated in the abstract reviewed · 1 Nov 2026 · computer science

Discovery

From Single-Agent Self-Discovery to Multi-Agent Adaptation: Scaling Laws of Partner Feedback and Convention Rigidity in Meta Reinforcement Learning Algorithms

Source check: Source checked. Not independent scientific confirmation.

Finding: Across more than 1,800 runs in simulated collective-robotics and coordination tasks, partner swaps revealed that apparent team-return gains could conceal lower focal-agent performance. Results also showed sensitivity to reciprocal partner feedback; they are simulation findings.

Technology: Disco103 and OPEN meta-learned reinforcement-learning update rules

Publication: Preprint · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.

AI contribution
Disco103 and OPEN meta-learned reinforcement-learning update rules
Test and result
Across 1,800+ simulated runs, four-cell partner swaps found an apparent +1.066 team-return gain alongside a −0.119 focal-agent loss in a reported setting.
Limitation / still unknown
The results concern specified VMAS and Overcooked simulations; they do not establish performance in physical multi-agent systems.
Evidence sources and annotation context

Evidence context (descriptions follow the cited sources; test type is a catalog classification):

Evidence sources: https://arxiv.org/abs/2611.01742

Source checked · editorial details

The source page was checked against this entry. Independent confirmation is not assessed by this label.

Source type: paper. Paper: linked separately.

Review date: 22 Nov 2026. Review scope: Primary result-bearing material for the stated method, result and evaluation, with bibliographic version information..

No follow-up recorded; this is not a claim that no follow-up exists. Follow-up review scope and date are not recorded. No source-supported relationships recorded.

Editorial assessment: Impact 3/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.

Open in the full catalog for filters, charts and related discoveries. Browse computer science discoveries.