Metraipolis › archive › dc_f5a2a3370145

Perturbation-Based Data Selection for Medical Question Answering — arXiv:2610.33440

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.33440
License (as archived)
CC0-1.0
Archived
2026-10-27T08:11:19.849Z
SHA-256 of the archived text
3ffa04fbce74f61b14e4c9ea539a9c605cbd1face7f0f998c8943c2c0a61a2ec
Stable URL of this page
https://metraipolis.com/documento/dc_f5a2a3370145

How to cite

Perturbation-Based Data Selection for Medical Question Answering — arXiv:2610.33440 — arXiv (en), https://arxiv.org/abs/2610.33440 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_f5a2a3370145 (Metraipolis archive dc_f5a2a3370145, sha256 3ffa04fbce74f61b14e4c9ea539a9c605cbd1face7f0f998c8943c2c0a61a2ec)
BibTeX
@misc{dcf5a2a3370145,
  title = {Perturbation-Based Data Selection for Medical Question Answering — arXiv:2610.33440},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.33440},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_f5a2a3370145; sha256 3ffa04fbce74f61b14e4c9ea539a9c605cbd1face7f0f998c8943c2c0a61a2ec},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.33440. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Selecting a compact training subset is an important step in adapting medical language models from existing question-answering banks. We investigate perturbation-based data selection using two signals derived from Gaussian perturbations of the final-block representation: prediction-change frequency and non-flipping margin contraction. After a shared entropy prefilter, Representation-Sensitivity Prioritization (RSP) combines the signals using automatic root-mean-square scaling, while RSP-Drop ranks candidates by contraction alone. Both selectors use only questions and answer options during selection. We evaluate them with MedGemma-4B on five medical QA datasets at retained-set budgets of 64, 128, and 256 examples. At the primary budget of 128, RSP and RSP-Drop obtain dataset-averaged Macro-F1 scores of 47.55 and 48.28, respectively, compared with 45.11 for entropy-pool-matched random selection. Their relative performance varies across datasets and budgets. A selected-set case study identifies contrasting clean-margin profiles among the examples exchanged by the two rules, while a full-pool ablation shows that entropy prefiltering is not uniformly beneficial. These retrospective results characterize the empirical utility and limitations of perturbation-based selection under a fixed adaptation protocol. Evaluation uses one backbone and benchmarks with documented training-family overlap; it does not establish clinical effectiveness.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.CL, nuovi lavori del giorno).
Categorie arXiv: Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2610.33440 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-27T07:56:00.193Z dalla pagina https://arxiv.org/list/cs.CL/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).