Model card
BRIDO.
BRIDO Authorsopen-sourceBART-based with democratic contrastive learning for factual consistency
BRIDO: Bringing Democratic Order to Abstractive Summarization. arXiv:2502.18342 (Feb 2025). Extends BRIO with a democratic ordering objective that prioritizes factual consistency, trading slight ROUGE drops for 3.82% improvement on G-Eval consistency vs BRIO.
§ 01 · Benchmarks
Every benchmark BRIDO has a recorded score for.
| # | Benchmark | Area · Task | Metric | Value | Rank | Date | Source |
|---|---|---|---|---|---|---|---|
| 01 | cnn-/-daily-mail | Computer Vision · Optical Character Recognition | rouge-2 | 22.9% | #4 | 2025-02-25 | source ↗ |
| 02 | cnn-/-daily-mail | Computer Vision · Optical Character Recognition | rouge-l | 42.5% | #4 | 2025-02-25 | source ↗ |
| 03 | cnn-/-daily-mail | Computer Vision · Optical Character Recognition | rouge-1 | 45.8% | #5 | 2025-02-25 | source ↗ |
Rank column shows this model’s position vs all other models scored on the same benchmark + metric (competitors after the slash). #1 in red means current SOTA. Sorted by rank, then newest result.
§ 03 · Papers
1 paper with results for BRIDO.
- 2025-02-25· Natural Language Processing· 3 results
BRIDO: Bringing Democratic Order to Abstractive Summarization
§ 05 · Sources & freshness
Where these numbers come from.
arxiv
3
results
3 of 3 rows marked verified.