Two questions regarding practitioners' use of patent embeddings arise: (i) Does one fine-tuning recipe suffice for all downstream applications? (ii) Is fine-tuning on one patent landscape sufficient for downstream application on other landscapes? By evaluating 22 pre-trained embedding models (ranging from 22M to 12B parameters) on three tasks -- information retrieval, classification, and clustering -- on 113,148 WIPO patents for assistive technology (46,069 citation queries) and on an external DAPFAM dataset, we find that two results cast doubt on the prevailing wisdom. (i) The optimal fine-tuning recipe depends on the downstream task: cross-sectional alignment (recipe R3) provides the largest improvements to retrieval performance (+7.1% nDCG@10), whereas a combined signal recipe (recipe R4) is better suited to classification (+7.1 F1) and clustering (+10.9 V-measure); a matched data control confirms that differences in training dataset size are not a contributing factor. (ii) Single-landscape fine-tuning hampers cross-landscape information retrieval: fine-tuning on one landscape significantly degrades cross-domain retrieval for 5 of 8 model-recipe combinations on the DAPFAM corpus, with the stronger zero-shot models suffering most. While within-family scaling is consistent (Qwen3 0.6B->4B->8B; Llama-Nemotron 1B->8B), cross-family scaling is erratic; the 12B KaLM-Gemma3 is ranked 8th on TAC retrieval performance, following prefix modification. Title+Abstract+Claims is the ubiquitous best text view, and all models suffer from a 55-65% gap between IN and OUT-of-domain performance which cannot be mitigated by hybrid BM25-dense fusion. Code and evaluation framework are publicly available.

翻译：实践者在专利嵌入应用中面临两个关键问题：(i) 是否存在适用于所有下游应用的统一微调方案？(ii) 在单个专利领域进行微调是否足以支撑其他领域的下游应用？通过对113,148项辅助技术领域世界知识产权组织（WIPO）专利（含46,069条引文查询）及外部DAPFAM数据集，评估22个预训练嵌入模型（参数量从22M到12B）在信息检索、分类和聚类三项任务上的表现，我们发现两项结果对现有认知提出质疑：(i) 最优微调方案取决于下游任务——横截面对齐方案（方案R3）对检索性能提升最大（nDCG@10提升7.1%），而组合信号方案（方案R4）更适合分类（F1值提升7.1）和聚类（V-measure提升10.9）；匹配数据对照实验证实训练集规模差异并非影响因素。(ii) 单领域微调会损害跨领域信息检索：在DAPFAM语料库中，8个模型-方案组合中有5个因单领域微调导致跨域检索性能显著下降，其中零样本能力较强的模型受损最大。虽然同家族模型具备一致的规模缩放规律（Qwen3 0.6B→4B→8B；Llama-Nemotron 1B→8B），但跨家族缩放则呈现不规律性；参数量12B的KaLM-Gemma3经前缀调整后，在TAC检索性能中仅排名第8。标题+摘要+权利要求书构成普适最优文本视图，所有模型在域内与域外性能间存在55-65%的差距，且无法通过混合BM25-密集融合技术弥补。代码与评估框架已公开提供。