arXiv：cs.CL· Ilya Chekin (BroutonLab), Vyacheslav Malyugin (BroutonLab), Vladimir Chirkov (BroutonLab), Mikhail Yurushkin (Curately)·· 4 小时前AI 评分34
通过蒸馏生产环境 LLM 信号构建可解释的简历-职位匹配特征表示
Building Interpretable Feature Representations for Resume-Vacancy Matching by Distilling Production LLM Signals
AI 导读
研究者提出一种两阶段简历-职位匹配方法：先用基于 LLM 的标注器生成可解释的匹配维度标签，再蒸馏出可在 CPU 上运行的特征双编码器（LoRA 适配的嵌入骨干加各维度紧凑预测头）处理全部在线请求。模型基于 168,772 对已标注的职位-简历数据训练，在 927 条招聘人员记录的生产反馈数据上，部署的学生模型与记录决策一致率达 95.79%。该结果属于运营性非盲一致性，而非独立人工评估。
正文
Abstract:Matching candidates to vacancies is central to recruitment, and a recruiter needs to see why a candidate fits, not only a single opaque relevance score. We provide this evidence as named, interpretable matching dimensions recruiters can act on - eight in our current deployment. We propose a two-part approach. The first is an LLM-based labeler whose prompts and feature definitions were refined from recruiter feedback while it served as an earlier production matching stage. In the current architecture, it is used only for offline labeling and is not called on online requests. The second is a feature bi-encoder distilled from it: a LoRA-adapted embedding backbone with compact per-dimension heads that runs on CPU and serves all online requests. Both parts keep improving: prompts are revised as feedback arrives, and the bi-encoder is retrained on the updated labels. The model is trained on 168,772 labeled vacancy-resume pairs (17,921 vacancies and 180,030 resumes). Recruiters using the service can confirm or revise surfaced feature predictions. On 927 recruiter-recorded values from this selected production-feedback subset, the deployed student agrees with the recorded decisions in 888 cases (95.79%). This is operational, non-blinded agreement rather than an independent human evaluation.
| Comments: | Accepted to EMNLP 2026; 13 pages, 4 figures, 7 tables |
| Subjects: | Computation and Language (cs.CL) |
| Cite as: | arXiv:2610.03112 [cs.CL] |
| (or arXiv:2610.03112v1 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2610.03112 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Ilya Chekin [view email]
[v1]
Fri, 2 Oct 2026 10:30:32 UTC (40 KB)
来源：arXiv：cs.CL · arxiv.org