Meeting Planning on Natural Plan
95AccuracyMultiGA (GPT-5 seed)
Evaluation results
| Method | Links | ||
|---|---|---|---|
| MultiGA (GPT-5 seed)Eval Model=Qwen3, Shot count=5-shot2025.11 | 95 | 81,770 | |
| MultiGA (Ensemble GA)Eval Model=Qwen3, Shot count=5-shot2025.11 | 88 | 71,499 | |
| GPT-5Shot count=N-shot2025.11 | 80.6 | 9,873 | |
| Qwen3Shot count=0-shot2025.11 | 79.6 | 7,877 | |
| GPT-5Shot count=0-shot2025.11 | 77.5 | 10,997 | |
| MultiGA (Llama-4 seed)Eval Model=Qwen3, Shot count=5-shot2025.11 | 75 | 107,670 | |
| MultiGA (Ensemble GB)Eval Model=GPT-5, Shot count=5-shot2025.11 | 69 | 92,784 | |
| Qwen3Shot count=N-shot2025.11 | 62.2 | 6,345 | |
| MultiGA (Gemma-2 seed)Eval Model=Qwen3, Shot count=5-shot2025.11 | 32 | 52,673 | |
| MultiGA (Phi-4-Mini seed)Eval Model=Qwen3, Shot count=5-shot2025.11 | 0 | — |