Journals of Accelerator Conferences Website (JACoW)
JACoW
is a publisher in Geneva, Switzerland that publishes the
proceedings of accelerator conferences held around the world
by an international collaboration of editors.
| URL | https://doi.org/10.18429/JACoW-IPAC2025-THPM023 |
|---|---|
| Title | The journey towards a specialized text embedding model for accelerator physics |
| Authors |
|
| Abstract | We present PhysBERT and AccPhysBERT, specialized sentence-embedding models trained on 1.2 million arXiv physics papers and fine-tuned for accelerator physics, respectively. Evaluation across retrieval, clustering, and similarity tasks shows gains of up to 12\% over general-purpose models for physics corpora and 18\% for accelerator-specific tasks. Applications include semantic reviewer–paper matching, Retrieval-Augmented Generation for control-room logbooks, and rapid sub-domain adaptation. We analyze key design choices—data curation, masking objectives, and contrastive fine-tuning—and outline strategies for continual adaptation, providing a blueprint for domain-specific embeddings in the physical sciences. |
| Paper | download: THPM023.pdf |
| Cite | BibTeX, LaTeX, Text/Word, RIS, EndNote |
| Conference | 16th International Particle Accelerator Conference |
| Series | |
| Location | Taipei, Taiwan |
| Date | 01-06 Jun 2025 |
| Publisher | JACoW Publishing, Geneva, Switzerland |
| Editorial Board | * Ming-Chyuan Lin - National Synchrotron Radiation Research Center * Yoichi Sato - KEK * Jui-Che Huang - National Synchrotron Radiation Research Center * David Button - Australian Nuclear Science and Technology Organisation * Ping-Shun Chuang - National Synchrotron Radiation Research Center |
| Online ISBN | 978-3-95450-248-6 |
| Online ISSN | 2673-5490 |
| Received | 30 May 2025 |
| Revised | 05 June 2025 |
| Accepted | 05 June 2025 |
| Issued | 05 November 2025 |
| DOI | 10.18429/JACoW-IPAC2025-THPM023 |
| Pages | 2730-2732 |
| Copyright | Published by JACoW Publishing under the terms of the Creative Commons Attribution 4.0 license. Any further distribution of this work must maintain attribution to the author(s), the published article's title, publisher, and DOI. |