mlsanityData to 14 Oct 2026

Building Interpretable Feature Representations for Resume-Vacancy Matching by Distilling Production LLM Signals

arXiv 2610.03112 ↗First posted 2 Oct 2026cs.CLcited by 0

Not on the map. Why?

Abstract (arXiv)

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.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 19 of its 25 references that are in the map. First posted on arXiv 2 Oct 2026.

Builds on

19 of 25 references.

References · 19

  1. 2015Distilling the Knowledge in a Neural Networkcited by 10,324
  2. 2019Sentence-BERT: Sentence Embeddings using Siamese BERT-Networkscited by 9,170
  3. 2020Dense Passage Retrieval for Open-Domain Question Answeringcited by 3,824
  4. 2025Qwen3 Embedding: Advancing Text Embedding and Reranking Through Foundation Modelscited by 1,261
  5. 2023ChatGPT Outperforms Crowd-Workers for Text-Annotation Taskscited by 870
  6. 2023Large Language Modelscited by 780
  7. 2019Passage Re-ranking with BERTcited by 645
  8. 2017Supervised Learning of Universal Sentence Representations from Natural Language Inference Datacited by 432
  9. 2016Enhanced LSTM for Natural Language Inferencecited by 200
  10. 2021RocketQAv2: A Joint Training Method for Dense Passage Retrieval and Passage Re-rankingcited by 176
  11. 2019Distilling Task-Specific Knowledge from BERT into Simple Neural Networkscited by 153
  12. 2020Improving Efficient Neural Ranking Models with Cross-Architecture Knowledge Distillationcited by 145
7 more
  1. 2020Augmented SBERT: Data Augmentation Method for Improving Bi-Encoders for Pairwise Sentence Scoring Taskscited by 138
  2. 2015Natural Language Inference by Tree-Based Convolution and Heuristic Matchingcited by 43
  3. 2020Learning to Match Jobs with Resumes from Sparse Interaction Data using Multi-View Co-Teaching Networkcited by 25
  4. 2025Scaling Down, Serving Fast: Compressing and Deploying Efficient LLMs for Recommendation Systemscited by 2
  5. 2025Enhancing Talent Search Ranking with Role-Aware Expert Mixtures and LLM-based Fine-Grained Job Descriptionscited by 2
  6. 2026ConFit v3: Improving Resume-Job Matching with LLM-based Re-Rankingcited by 2
  7. 2025LANTERN: Scalable Distillation of Large Language Models for Job-Person Fit and Explanationcited by 1

Built on by

Not cited yet.