Science Question Answering on GPQA-D (Acc, Tok)
68.5AccuracyPACT
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| PACTBase Model=Qwen3-32B, Pipeline Setting=four-agent sequential pipeline (Setting B)2026.06 | 68.5 | 16,271 | |
| Multi-Agent DebateBase Model=Qwen3-32B, Pipeline Setting=four-agent sequential pipeline (Setting B)2026.06 | 66.4 | 77,888 | |
| CoABase Model=Qwen3-32B, Pipeline Setting=four-agent sequential pipeline (Setting B)2026.06 | 65.7 | 18,535 | |
| TextMASBase Model=Qwen3-32B, Pipeline Setting=four-agent sequential pipeline (Setting B)2026.06 | 65.7 | 21,428 | |
| PACTBase Model=Qwen3-14B, Pipeline Setting=four-agent sequential pipeline (Setting B)2026.06 | 62.5 | 15,935 | |
| Multi-Agent DebateBase Model=Qwen3-14B, Pipeline Setting=four-agent sequential pipeline (Setting B)2026.06 | 61.8 | 81,878 | |
| CoABase Model=Qwen3-14B, Pipeline Setting=four-agent sequential pipeline (Setting B)2026.06 | 61.6 | 18,789 | |
| TextMASBase Model=Qwen3-14B, Pipeline Setting=four-agent sequential pipeline (Setting B)2026.06 | 61 | 16,366 | |
| PACTBase Model=Qwen3-8B, Pipeline Setting=four-agent sequential pipeline (Setting B)2026.06 | 60.4 | 19,960 | |
| CoABase Model=Qwen3-8B, Pipeline Setting=four-agent sequential pipeline (Setting B)2026.06 | 59.5 | 22,563 | |
| TextMASBase Model=Qwen3-8B, Pipeline Setting=four-agent sequential pipeline (Setting B)2026.06 | 59 | 20,357 | |
| Multi-Agent DebateBase Model=Qwen3-8B, Pipeline Setting=four-agent sequential pipeline (Setting B)2026.06 | 57.7 | 85,351 |