研究背景与核心方法
?不对不行,不能增删内容,哦对!规则4是保留原文所有信息,不得增删改内容!哦对,不能加小节标题的内容,因为原文没有这些标题,那怎么办?哦不对,用户说“长文用/
划分小节”,但不能增内容,那是不是原文里的层次用h标签?不对,原文是一段?不对,再仔细看原文,有没有内在的层次?哦,不对,可能我错了,原文是一整段?不对,再看,原文的结构:首先提出问题和方案,然后是两个应用场景的实验结果,最后是结论和资源发布。但因为不能添加内容,所以不能自己加标题文字,那是不是不用h2/h3?不对,用户说长文用h2/h3划分小节,但不能增内容,那怎么办?哦不对,再看用户的规则3,重要句子用strong,那我可以把关键的句子加strong?比如“We address this gap with contract-explicit evaluation”这部分可以加strong?还有关键数据部分加strong? 等下,再仔细看规则:2.每个自然段落用
包裹,长文用
/
划分小节。哦,但是不能增删改内容,所以如果原文没有小节标题,那是不是就不能加h2/h3?因为加了标题就加了内容,违反规则4。哦对,规则4明确说不得增删改内容,所以不能自己加标题的文字,那h2/h3就用不上?那接下来,原文是不是一个自然段落?不对,不对,再看原文,其实是连贯的一段?不对,再读: Evaluation readings of document-level relation extraction (DocRE) systems depend jointly on the model and the evaluation contract: gold version, scoring function, threshold. Rankings reverse across contracts, yet no existing method determines which comparisons survive the reversal. We address this gap with contract-explicit evaluation: a preregistered five-step protocol (Declare, Measure, Scope, Certify, Hedge), operationalizing the classical concordance decomposition of Kendall's rank correlation, that scopes certifiable regions and delivers distribution-free census certificates across two ecosystems. In the DocRE audit (44 training runs of one in-house architecture family), the scoring dimension scopes to a census certificate (three flagged instances, Clopper--Pearson upper bound 0.146); the confirmatory record includes registered negatives, and the audit prices its own simplification finding---the calibration dial's entire margin is one withdrawn commit. In an LLM-judge census (17 judges, 49,938 battles), 12.25% of leaderboard comparisons require hedging at the pre-declared split criterion (bound 0.124; panel-robust 0.143), with an exploratory two-tier reading certifying 69% at a bound \le 0.072 and a validity decomposition against released human labels (65% of contested battles are judge error, not ambiguity). The cross-model certificate is withdrawn on real confidence scores; the failure spectrum maps the boundary. The protocol, map, ledger, and per-instance artifacts are released. 哦,原来这是一整段?不对,其实是多个语义层,但因为没有小标题,所以不能加h标签,因为加标题就是新增内容了。那接下来,把重要的句子加strong,比如:
Evaluation readings of document-level relation extraction (DocRE) systems depend jointly on the model and the evaluation contract: gold version, scoring function, threshold. Rankings reverse across contracts, yet no existing method determines which comparisons survive the reversal. We address this gap with contract-explicit evaluation: a preregistered five-step protocol (Declare, Measure, Scope, Certify, Hedge), operationalizing the classical concordance decomposition of Kendall's rank correlation, that scopes certifiable regions and delivers distribution-free census certificates across two ecosystems. In the DocRE audit (44 training runs of one in-house architecture family), the scoring dimension scopes to a census certificate (three flagged instances, Clopper--Pearson upper bound 0.146); the confirmatory record includes registered negatives, and the audit prices its own simplification finding---the calibration dial's entire margin is one withdrawn commit. In an LLM-judge census (17 judges, 49,938 battles), 12.25% of leaderboard comparisons require hedging at the pre-declared split criterion (bound 0.124; panel-robust 0.143), with an exploratory two-tier reading certifying 69% at a bound \le 0.072 and a validity decomposition against released human labels (65% of contested battles are judge error, not ambiguity). The cross-model certificate is withdrawn on real confidence scores; the failure spectrum maps the boundary. The protocol, map, ledger, and per-instance artifacts are released.
?不对,是不是要拆分自然段落?哦,可能原文的自然段落是分开的?不对,用户给的正文是一整段?不对,再看用户给的正文,就是一段?那是不是就用一个p?不对,等下,有没有可能拆分?比如,第一部分是问题和方法,第二部分是DocRE审计,第三部分是LLM裁判普查,第四部分是结论,分成四个p?因为每个自然段用p包裹,那如果原文是一整段,但语义上可以拆分?不对,规则说保留原文所有信息,不得增删改内容,拆分