arXiv cs.CL· Jorge Ram\'on Fonseca Cacho·· 11 小时前AI 评分27
EditVeto：面向 LLM 后 OCR 纠错的改写器无关编辑门控
Rewrite, Then Veto: Rewriter-Agnostic Edit Gating for LLM-Based Post-OCR Correction
AI 导读
研究者提出 EditVeto，在 LLM 全文改写后加一道 0.6B 参数的门控，按字符对齐把改写差异切块并逐块决定保留或回退。在 HIPE-OCRepair-2026 八个测试集上，该门控对五个改写器均有提升，其中两个未参与训练，加权 cMER 下降 0.0026-0.0044，偏好分数上升 0.15-0.25。
正文 · 原文
Abstract:LLMs can repair noisy OCR transcripts of historical documents, but they also change text that was already correct. We propose EditVeto, a veto gate placed after a full-context LLM rewrite. The rewrite is character-aligned to the OCR input, its differences are split into blocks, and a 0.6B-parameter gate keeps or reverts each block. An oracle analysis motivates the design: given the true error locations, per-span generation is worse than no correction, whereas vetoing a full rewrite's edits helps. Trained on rewrites by two Qwen3 models and evaluated, frozen, on the eight HIPE-OCRepair-2026 test sets, the gate improves all five rewriters tested, two of them unseen in training: weighted cMER falls by 0.0026-0.0044 and the preference score rises by 0.15-0.25 (paired bootstrap, p ~ 0.0002 each). On published competition outputs it lowers cMER for three of four teams but raises it for the winning run (0.0050 to 0.0065). In a second, disclosed test look, a retrained gate still improves LoRA-tuned rewriters (cMER -0.0011 to -0.0021), and an ensemble of their gated edits reaches a cMER of 0.0069 and a preference score of 0.928, mostly through fine-tuning and pooling rather than gating. Built only on public models and data, it is behind the winner (0.0050) and tied with the second-ranked run (0.0071), both from the only team using domain pre-training, and ahead of all other published runs. A third, secondary look shows the unchanged gate improving zero-shot Qwen3.8-27B (cMER 0.0114 to 0.0108) but only the preference score of its LoRA-tuned version.
| Comments: | 26 pages, 3 figures, 12 tables. Code: this https URL models: this https URL |
| Subjects: | Computation and Language (cs.CL); Computer Vision and Pattern Recognition (cs.CV) |
| Cite as: | arXiv:2611.05539 [cs.CL] |
| (or arXiv:2611.05539v1 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2611.05539 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Jorge Fonseca Cacho [view email]
[v1]
Tue, 3 Nov 2026 19:31:36 UTC (277 KB)
来源：arXiv cs.CL · arxiv.org