Summarization on CNN/DM (BERTScore and QAFactEval)
0.89BERTScorePEGASUS
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| PEGASUS2026.02 | 0.89 | 0.79 | — | |
| MatchSum2026.02 | 0.88 | 0.78 | — | |
| GloSA-sum2026.02 | 0.88 | 0.78 | — | |
| BERTSum2026.02 | 0.87 | 0.77 | — | |
| BIGBIRD2026.02 | 0.87 | 0.78 | — | |
| BART2026.02 | 0.86 | 0.77 | — | |
| LEAD-32026.02 | 0.83 | 0.68 | — | |
| TextRank2026.02 | 0.73 | 0.58 | — | |
| TaskFT-FullFine-tuning=Task-adaptive, Visibility Policy=Full2026.05 | — | — | 27.4 | |
| TaskFT-SPEED-20+BoSFine-tuning=Task-adaptive, K=20, BoS=true2026.05 | — | — | 28.1 | |
| TaskFT-SPEED-24+BoSFine-tuning=Task-adaptive, K=24, BoS=true2026.05 | — | — | 27.5 | |
| TaskFT-SPEED-28+BoSFine-tuning=Task-adaptive, K=28, BoS=true2026.05 | — | — | 27.8 |