List your company

Question Answering on BBQ (Accuracy and Token Usage)

99.5AccuracyMultiGA (Llama-4 seed)
86.91690.18393.4596.717Nov 21, 2025

Evaluation results

MethodLinks
MultiGA (Llama-4 seed)Eval Model=Qwen32025.1199.523,596
MultiGA (Ensemble GA)Eval Model=Qwen32025.1196.636,574
Qwen3Shot count=0-shot2025.1196.1168
MultiGA (GPT-5 seed)Eval Model=Qwen32025.1196.115,355
MultiGA (Gemma-2 seed)Eval Model=Qwen32025.1196.19,523
GPT-5Shot count=0-shot2025.1195.1128
MultiGA (Phi-4-Mini seed)Eval Model=Qwen32025.1192.333,629
MultiGA (Ensemble GB)Eval Model=GPT-52025.1187.437,723