Token Cost on MMLU-Pro (Medical Reasoning)
1,350Token Cost (tokens/question)DyLAN
Evaluation Results
| Method | Links | |
|---|---|---|
| DyLANModel=GPT-4o, Framework=DyLAN2025.08 | 1,350 | |
| ReConcileModel=GPT-4o, Framework=ReConcile2025.08 | 1,925 | |
| TMA-AllComponModel=MedGemma-4B, Framework=TMA-AllCompon2025.08 | 3,427 | |
| TMA-AllComponModel=Gemma-3-4B, Framework=TMA-AllCompon2025.08 | 3,427 | |
| TMA-AllComponModel=GPT-4o, Framework=TMA-AllCompon2025.08 | 3,427 | |
| ReConcileModel=Gemma-3-4B, Framework=ReConcile2025.08 | 6,271 | |
| DyLANModel=Gemma-3-4B, Framework=DyLAN2025.08 | 7,977 | |
| MedAgentsModel=Gemma-3-4B, Framework=MedAgents2025.08 | 10,362 | |
| MedAgentsModel=GPT-4o, Framework=MedAgents2025.08 | 10,362 | |
| MDAgentsModel=Gemma-3-4B, Framework=MDAgents2025.08 | 21,057 | |
| MDAgentsModel=GPT-4o, Framework=MDAgents2025.08 | 21,057 |