Software Engineering Task Solving on SWE-bench
73.41ASAgentSociety
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| AgentSocietyAgents=6, Domains=5, Model Configuration=Strong/ Generalist2026.05 | 73.41 | 70.24 | 3.2 | 73.81 | |
| AgentSocietyAgents=6, Domains=5, Model Configuration=Weaker/ Complementary2026.05 | 56.35 | 50.79 | 5.6 | 58.73 |