Multi-Agent Collaboration on SRDD
76.35CompletenessChatDev (Llama)
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| ChatDev (Llama)Backbone=Llama, Fine-tuning (FT)=true2026.04 | 76.35 | 87.16 | 39.37 | 67.63 | 0 | 84.31 | |
| ChatDev (Qwen)Backbone=Qwen, Fine-tuning (FT)=true2026.04 | 72.72 | 95.61 | 38.94 | 69.09 | 15.54 | 88.17 | |
| ChatDev (Llama)Backbone=Llama, Fine-tuning (FT)=false2026.04 | 71.06 | 82.43 | 32.18 | 61.74 | 62.25 | 50.03 | |
| ChatDev (Qwen)Backbone=Qwen, Fine-tuning (FT)=false2026.04 | 70.02 | 97.13 | 35.93 | 67.69 | 81.33 | 52.78 |