JEV 生态评测：13 个答案验证器同台竞技
原标题：Inside the JEV Ecosystem: 13 Answer Verifiers on One Test Set
AI 摘要
Hugging Face 博客发布了一项针对答案验证（answer verification）系统的独立评测，在 2,018 条统一标注的测试集上对比了 13 个系统，并公开了评分代码。结果显示仅三个系统 AUC 超过 0.70，最强的 ZTC (397B) 与 JEV 差距仅 0.0014，统计上无法区分；而仅依赖答案长度和格式的基线达到 0.7036，高于 13 个系统中的 8 个。在 20% 重试预算的 agent 端到端测试中，ZTC 门控带来 +1.34 个百分点提升，JEV 则为 -0.07 个百分点，说明验证器的精度而非召回率决定其实际价值。
正文节选
Every answer-verification vendor publishes a benchmark, and every one of them wins it. We ran thirteen systems on a single test set of 2,018 items with identical labels, published the scoring code, and put the results side by side. Typed Decision Leaderboard → https://huggingface.co/spaces/mayafree/typed-decision-leaderboard The short version: only three systems clear 0.70, the strongest two cannot be separated statistically, and a baseline that looks at nothing but answer length and formatting