AI 中文总结
研究专利权利要求生成中自回归解码器的局限，提出SPG方法，通过指针头选择父权利要求，结合梯度与正则化器重塑解码器表示，经两阶段训练，在HUPD-DCG上恢复大量黄金父链接并提升先行一致性。
AI 中文摘要
自回归解码器生成的是扁平令牌序列，无法在输出段之间强制实施层次约束，这在专利权利要求生成中是个严重问题，因为权利要求集构成依赖森林，范围随深度单调缩小。拓扑和内容相互依赖，事后解析和语法约束解码都不足。我们提出了SPG（结构感知专利生成），在自回归过程中预测拓扑。指针头选择每个从属权利要求的父权利要求，其梯度与深度自适应范围正则化器一起在训练期间重塑共享解码器的表示。第二阶段对自生成的有缺陷候选者应用违反加权偏好目标，提供授权专利语料库中缺乏的负信号。在HUPD-DCG上，Llama-3-8B-Instruct上的SPG恢复了79.0%的黄金父链接，提高了先行一致性，专家评估也证实了这些收益。
英文摘要
Autoregressive decoders emit flat token sequences and cannot enforce hierarchical constraints across output segments, a limitation that becomes acute in patent claim generation, where a claim set forms a dependency forest whose scope must narrow monotonically with depth. Topology and content are mutually dependent: a dependent claim's wording must reflect its parent's scope, yet the parent must be chosen before that wording exists, so neither post-hoc parsing nor grammar-constrained decoding suffices. We propose SPG (Structure-aware Patent Generation), which predicts topology inside the autoregressive pass. A pointer head selects each dependent claim's parent, and its gradients, together with a depth-adaptive scope regularizer, reshape the shared decoder's representations during training. A second stage then applies a violation-weighted preference objective over self-generated deficient candidates, supplying the negative signal that granted-patent corpora lack. On HUPD-DCG, SPG on Llama-3-8B-Instruct recovers 79.0\% of gold parent links, a quantity its training reward never supervises, and raises antecedent consistency from 0.292 to 0.478 over a supervised baseline of equal scale, with expert evaluation corroborating these gains.
