SOURCE-LINKED INTELLIGENCE

Don't Count the Edits, Judge by the Outcome Alone: Reward-Based Evaluation for Grammatical Error Correction

arXiv · AI, language, vision and robotics · article · Sep 14, 2026 · UTC

Grammatical error correction (GEC) evaluation has traditionally relied on reference or edit overlap, which can penalize valid rewrites that differ from gold corrections. Reference-free metrics reduce this dependence, but evaluating whether a fluent output is a valid correction of the source remains challenging. We propose SURE, a source-conditioned reward evaluator trained on within-source preferences spanning minimal-edit and rewrite-oriented corrections. SURE jointly learns an overall reward with criteria-level supervision for grammaticality, faithfulness, and fluency, together with span-lev

Read original source ↗ Open in workspace

recordType

paper

region

Global