Metraipolis › archive › dc_addcbe2c9dac

Rank-Aware Differentially Private Release of Listwise Preferences for LLM Alignment — arXiv:2610.37682

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.37682
License (as archived)
CC0-1.0
Archived
2026-10-31T06:10:56.035Z
SHA-256 of the archived text
833f638c8f9fc09fdfda1343137dbffa632b5e7dc411dc3fdea6e01ccf92aabe
Stable URL of this page
https://metraipolis.com/documento/dc_addcbe2c9dac

How to cite

Rank-Aware Differentially Private Release of Listwise Preferences for LLM Alignment — arXiv:2610.37682 — arXiv (en), https://arxiv.org/abs/2610.37682 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_addcbe2c9dac (Metraipolis archive dc_addcbe2c9dac, sha256 833f638c8f9fc09fdfda1343137dbffa632b5e7dc411dc3fdea6e01ccf92aabe)
BibTeX
@misc{dcaddcbe2c9dac,
  title = {Rank-Aware Differentially Private Release of Listwise Preferences for LLM Alignment — arXiv:2610.37682},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.37682},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_addcbe2c9dac; sha256 833f638c8f9fc09fdfda1343137dbffa632b5e7dc411dc3fdea6e01ccf92aabe},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.37682. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Listwise preference feedback offers richer supervision for large language model (LLMs) alignment than pairwise comparisons, but a full ranking also reveals more sensitive user preferences. Existing privacy-preserving alignment methods focus mainly on pairwise feedback, while listwise alignment under local differential privacy (LDP) remains unexplored. Extending LDP to this listwise domain raises a critical challenge: the high sensitivity of the full ranking space demands excessive noise, neutralizing the benefits of listwise supervision. To address this challenge, we analyze the interaction between supervision granularity and privacy noise in downstream training, and introduce a rank-aware exponential mechanism that privatizes listwise preference data into a low-sensitivity granularity suitable for downstream alignment. The mechanism leverages ranking information to sample a fixed-size binary partition, concentrating the noise near borderline items instead of perturbing all items uniformly as randomized response (RR) does. Empirical evaluations on downstream LLM alignment tasks show that our mechanism consistently outperforms existing LDP baselines at matched privacy budgets. The code is available at this https URL .

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Computation and Language (cs.CL) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2610.37682 — 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 2026-10-31T06:07:08.578Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).