Bench-MFG

Bench-MFG evaluates algorithms for learning stationary mean-field games across standardized discrete environments and generated instances.

unassessed

This page is a discovery lead. Nobody has yet assessed it against the catalogue contract, so it carries no disposition. Absence of evidence here is not evidence of staleness.
Categoryreasoning
Subcategorymulti-agent reinforcement learning
Metricexploitability
Directionlower_is_better
Unitscore
PublisherBench-MFG authors

What it measures

Bench-MFG evaluates learning methods for discrete-time, discrete-space stationary mean-field games. Its suite spans no-interaction, monotone, potential, and dynamics-coupled games, including randomly generated MF-Garnets instances.

Task format

Multi-agent reinforcement-learning environments with equilibrium or exploitability evaluation.

Models reporting this benchmark

No model card in ModelSpec reports this benchmark yet.

Data

This page as JSON · Edit on GitHub