Factual Text Generation on FactScore
0.7345AURCSelective Redaction
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Selective RedactionModel=GPT-OSS-20B2026.02 | 0.7345 | — | |
| Atom-wise Selective AbstractionModel=GPT-OSS-20B2026.02 | 0.6159 | 16.14 | |
| Selective RedactionModel=GPT-OSS-120B2026.02 | 0.613 | — | |
| Selective RedactionModel=QWEN3-30B-T2026.02 | 0.6098 | — | |
| Selective RedactionModel=QWEN3-30B-I2026.02 | 0.5784 | — | |
| Selective RedactionModel=Average2026.02 | 0.5751 | — | |
| Atom-wise Selective AbstractionModel=QWEN3-30B-T2026.02 | 0.5099 | 16.39 | |
| Atom-wise Selective AbstractionModel=QWEN3-30B-I2026.02 | 0.4903 | 15.23 | |
| Selective RedactionModel=QWEN3-235B-T2026.02 | 0.483 | — | |
| Atom-wise Selective AbstractionModel=Average2026.02 | 0.4749 | 17.43 | |
| Atom-wise Selective AbstractionModel=GPT-OSS-120B2026.02 | 0.443 | 27.73 | |
| Selective RedactionModel=LLAMA-3.3-70B-I2026.02 | 0.432 | — | |
| Atom-wise Selective AbstractionModel=LLAMA-3.3-70B-I2026.02 | 0.404 | 6.48 | |
| Atom-wise Selective AbstractionModel=QWEN3-235B-T2026.02 | 0.3862 | 20.04 |