Cooperative Multi-agent Problem Solving on Newsgroups (Easy) benchmark leaderboard · SOTA2 Research

Cooperative Multi-agent Problem Solving on Newsgroups (Easy)

1Detected ErrorMAS-Only

Evaluation Results

Method	Links
MAS-Only2026.02		1	2.4	11.47	100
MAS+LLM Judge2026.02		1	2.32	12.57	100
MAS+DIG2026.02		1	2.38	9.83	70
MAS-Only2026.02		2	2.11	21.98	100
MAS-Only2026.02		2	2.18	21.33	100
MAS+LLM Judge2026.02		4	2.19	28.83	70
MAS+LLM Judge2026.02		4	2.35	26.73	100
MAS+DIG2026.02		8	2.01	48.12	100
MAS+DIG2026.02		9	1.82	37.69	100