Scan-level report labelling on Herniation (test)
0.946Balanced AccuracySummary-Query
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| Summary-QuerySummary Model=Llama3, Query Model=Llama32024.10 | 0.946 | 0.071 | 0.99 | 0.931 | |
| Direct QueryQuery Model=GPT-42024.10 | 0.935 | — | — | 0.915 | |
| Direct QueryQuery Model=Llama32024.10 | 0.911 | 0.071 | 0.98 | 0.893 | |
| Direct QueryQuery Model=Zephyr2024.10 | 0.869 | 0.119 | 0.947 | 0.842 | |
| Summary-QuerySummary Model=Zephyr, Query Model=Zephyr2024.10 | 0.774 | 0.31 | 0.815 | 0.738 | |
| Summary-QuerySummary Model=Z-SFT, Query Model=Zephyr2024.10 | 0.768 | 0.214 | 0.871 | 0.73 |