Categories Companies Research

Loading the SOTA2 catalog…

Explore

Categories Companies Research

Contribute

Join SOTA2 Contact

SOTA2

© 2026 SOTA2

Build in Bellevue, USA

WizWand is now SOTA2 Research

Papers Benchmarks Datasets Tasks

Multi-Agent Path Finding research benchmarks · SOTA2 Research

Multi-Agent Path Finding

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

Single Policy Oracle (SPO)

1.28AUC Ratio (SPO/Baseline)

98

Feb 27, 2026

Single Policy Oracle (SPO)

1.47AUC Ratio (SPO / Best Baseline)

98

Feb 27, 2026

Single Policy Oracle (SPO)

1.21AUC Ratio

98

Feb 27, 2026

Large Maze 33x33 world size, 33.0% static obstacle rate

100Success Rate

20

May 14, 2026

Medium Warehouse 25x25 world size, 34.6% static obstacle rate

100Success Rate

20

May 14, 2026

Medium Room 23x23 world size, 33.5% static obstacle rate

100Success Rate

20

May 14, 2026

Medium Maze 25x25 world size, 32.8% static obstacle rate

100Success Rate

20

May 14, 2026

Small Random 10x10 world size, 17.5% static obstacle rate

96Success Rate

20

May 14, 2026

400x400x25 grid environment (4 temporal NFZs) (20 scenarios)

1Success Rate

20

Feb 26, 2026

400x400x25 grid environment 2 temporal NFZs (20 scenarios)

1Success Rate

20

Feb 26, 2026

400x400x25 grid environment 0 temporal NFZs (20 scenarios)

100Success Rate

20

Feb 26, 2026

Random Map 240x240, 0.3 density

98Success Rate (SR)

15

Feb 26, 2026

Random Map 120x120, 0.3 density

100Success Rate

15

Feb 26, 2026

random-32-32-20 map (test)

100Success Rate (SR)

12

Mar 20, 2026

3-agent Sum (evaluation set)

Qwen3-4B + GRPO + Ours

51.67Valid Rate

7

Jun 17, 2026

3-agent 10x10 (test)

Qwen3-4B + GRPO + Ours

32Valid Rate

7

Jun 17, 2026

3-agent 9x9 (test)

Qwen3-4B + GRPO + Ours

44Valid Rate

7

Jun 17, 2026

3-agent 8x8 (test)

Qwen3-4B + GRPO + Ours

42.67Valid Rate

7

Jun 17, 2026

3-agent 7x7 (evaluation)

Qwen3-4B + GRPO + Ours

46Valid Rate

7

Jun 17, 2026

3-agent 6x6 (evaluation set)

Qwen3-4B + GRPO + Ours

52.67Valid Rate

7

Jun 17, 2026

3-agent 5x5 (evaluation set)

Qwen3-4B + GRPO + Ours

62.67Valid Rate

7

Jun 17, 2026

3-agent 4x4 (test)

Qwen3-4B + GRPO + Ours

64.67Valid Rate

7

Jun 17, 2026

3-agent evaluation set (3x3)

Qwen3-4B + GRPO + Ours

68.67Valid Rate

7

Jun 17, 2026

MAPF-FrozenLake 4-agent Sum

Qwen3-4B + GRPO + LLM-as-Environment-Engineer

33.14Valid Rate

7

Jun 17, 2026

MAPF-FrozenLake 4-agent 10x10

Qwen3-4B + GRPO + LLM-as-Environment-Engineer

18.67Valid Rate

7

Jun 17, 2026

Page 1 of 3

PreviousNext

Popular datasets

Setup MAPF-FrozenLake 3-agent Random Map CL-CBS

Follow for updates

Join SOTA2 Contact