Selective Prediction on BioASQ
0.2744E-AURCSequence NLL
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Sequence NLLBase Model=Gemma 3 4B2026.03 | 0.2744 | 0.4657 | |
| SE ProbeBase Model=Qwen3 4B2026.03 | 0.2603 | 0.2458 | |
| Sequence NLLBase Model=Qwen3 4B2026.03 | 0.2513 | 0.2607 | |
| Sequence NLLBase Model=Llama 3.2 3B2026.03 | 0.2443 | 0.1987 | |
| PC ProbeBase Model=Llama 3.2 3B2026.03 | 0.2373 | 0.2632 | |
| Combined: PC + NLLBase Model=Llama 3.2 3B2026.03 | 0.2173 | 0.1666 | |
| SE ProbeBase Model=Llama 3.2 3B2026.03 | 0.2143 | 0.2942 | |
| Sequence NLLBase Model=Ministral 8B2026.03 | 0.2123 | 0.1876 | |
| Semantic EntropyBase Model=Qwen3 4B2026.03 | 0.2103 | 0.1977 | |
| Combined: PC + SEBase Model=Qwen3 4B2026.03 | 0.2023 | 0.3891 | |
| PC ProbeBase Model=Qwen3 4B2026.03 | 0.2013 | 0.3891 | |
| Combined: PC + SE ProbeBase Model=Qwen3 4B2026.03 | 0.1993 | 0.3871 | |
| Combined: PC + SE ProbeBase Model=Llama 3.2 3B2026.03 | 0.1942 | 0.1511 | |
| Combined: PC + NLLBase Model=Qwen3 4B2026.03 | 0.1923 | 0.3002 | |
| Combined: PC + SEBase Model=Llama 3.2 3B2026.03 | 0.1842 | 0.0952 | |
| Semantic EntropyBase Model=Llama 3.2 3B2026.03 | 0.1633 | 0.0731 | |
| PC ProbeBase Model=Gemma 3 4B2026.03 | 0.1623 | 0.1368 | |
| Combined: PC + NLLBase Model=Gemma 3 4B2026.03 | 0.1623 | 0.0953 | |
| Combined: PC + SE ProbeBase Model=Gemma 3 4B2026.03 | 0.1583 | 0.1084 | |
| Semantic EntropyBase Model=Gemma 3 4B2026.03 | 0.1523 | 0.0943 | |
| Combined: PC + SEBase Model=Gemma 3 4B2026.03 | 0.1482 | 0.0791 | |
| PC ProbeBase Model=Ministral 8B2026.03 | 0.1462 | 0.0732 | |
| SE ProbeBase Model=Gemma 3 4B2026.03 | 0.1423 | 0.1205 | |
| Combined: PC + NLLBase Model=Ministral 8B2026.03 | 0.1382 | 0.0701 | |
| SE ProbeBase Model=Ministral 8B2026.03 | 0.1333 | 0.1006 | |
| Semantic EntropyBase Model=Ministral 8B2026.03 | 0.1323 | 0.0713 | |
| Combined: PC + SE ProbeBase Model=Ministral 8B2026.03 | 0.1292 | 0.0691 | |
| Combined: PC + SEBase Model=Ministral 8B2026.03 | 0.1142 | 0.0611 |