Information Retrieval on SciFact (Faith., Hit@k, p95)
67Faithfulnessgpt-4.1-mini
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| gpt-4.1-miniScenario=Risk-first, k=5, Runs=32026.03 | 67 | 84 | 3,736 | |
| gpt-4.1-miniScenario=Cost-first, k=5, Runs=32026.03 | 63 | 84 | 3,324 | |
| gpt-4.1-miniScenario=SLA-first, k=5, Runs=32026.03 | 63 | 84 | 3,505 | |
| gpt-5.2Scenario=Cost-first, k=3, Runs=32026.03 | 62 | 82 | 6,224 | |
| gpt-5.2Scenario=SLA-first, k=3, Runs=32026.03 | 61 | 82 | 6,792 | |
| gpt-5.2Scenario=Risk-first, k=3, Runs=32026.03 | 60 | 82 | 7,072 | |
| gpt-4.1Scenario=Risk-first, k=5, Runs=32026.03 | 59 | 84 | 3,049 | |
| gpt-4.1Scenario=Cost-first, k=3, Runs=32026.03 | 58 | 82 | 2,933 | |
| gpt-4.1Scenario=Cost-first, k=5, Runs=32026.03 | 58 | 84 | 3,585 | |
| gpt-4.1Scenario=SLA-first, k=5, Runs=32026.03 | 58 | 84 | 3,200 | |
| gpt-4.1-miniScenario=Cost-first, k=3, Runs=32026.03 | 58 | 82 | 3,516 | |
| gpt-4.1-miniScenario=Risk-first, k=3, Runs=32026.03 | 58 | 82 | 3,744 | |
| gpt-4.1-miniScenario=SLA-first, k=3, Runs=32026.03 | 58 | 82 | 3,439 | |
| gpt-5.2Scenario=Cost-first, k=5, Runs=32026.03 | 58 | 84 | 6,474 | |
| gpt-5.2Scenario=SLA-first, k=5, Runs=32026.03 | 58 | 84 | 6,229 | |
| gpt-4.1Scenario=Risk-first, k=3, Runs=32026.03 | 56 | 82 | 3,304 | |
| gpt-5.2Scenario=Risk-first, k=5, Runs=32026.03 | 56 | 84 | 6,547 | |
| gpt-4.1Scenario=SLA-first, k=3, Runs=32026.03 | 55 | 82 | 3,154 |