Metraipolis › archive › dc_3dbe8a46a298

Auditing Multi-Proxy Explanations of Web User Preferences: Identification Boundaries and Abstention — arXiv:2611.04997

Source
arXiv (en)
Original
https://arxiv.org/abs/2611.04997
License (as archived)
CC0-1.0
Archived
2026-11-04T21:51:51.260Z
SHA-256 of the archived text
d3df485db52d858d023d981c33bf10877d714818a358863a91df9cefafcd987d
Stable URL of this page
https://metraipolis.com/documento/dc_3dbe8a46a298

How to cite

Auditing Multi-Proxy Explanations of Web User Preferences: Identification Boundaries and Abstention — arXiv:2611.04997 — arXiv (en), https://arxiv.org/abs/2611.04997 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_3dbe8a46a298 (Metraipolis archive dc_3dbe8a46a298, sha256 d3df485db52d858d023d981c33bf10877d714818a358863a91df9cefafcd987d)
BibTeX
@misc{dc3dbe8a46a298,
  title = {Auditing Multi-Proxy Explanations of Web User Preferences: Identification Boundaries and Abstention — arXiv:2611.04997},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.04997},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_3dbe8a46a298; sha256 d3df485db52d858d023d981c33bf10877d714818a358863a91df9cefafcd987d},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.04997. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Web systems infer user preferences from behavior, ratings, text, and LLM-generated profiles, yet explaining a proxy may reveal its measurement process rather than the preference of interest. We develop an audit framework distinguishing point identification, partial identification, and abstention relative to an explicit reference distribution. Under feature-dependent, arbitrarily signed proxy bias, a strict majority of uncontaminated proxy attributions lets the coordinatewise median recover latent-attribution support and direction and, with externally justified positive scales, magnitude. We prove this threshold sharp using response-valid indistinguishable worlds, characterize sharp minimax error and exact compatible-target sets under approximate cleanliness and scale uncertainty, and propagate clean-proxy attribution error. A frozen known-truth benchmark spans two response families, four proxy counts, all contamination counts, three bias strengths, and eleven methods, with a focal comparison confirmed on twenty unseen seeds. Under correct budgets, a direct attribution-tensor audit achieved 100% coverage, 95.40%-97.05% active-driver sign decisions, and no false signs. In a temporally separated Amazon study, median aggregation reduced comparator-relative conditional sign error for all four targets; six of eight prespecified target-metric contrasts had Bonferroni intervals excluding zero. A controlled real-covariate perturbation favored the median on the prespecified focal metric for all four targets. Compatible sets showed that the current proxy configuration rarely licensed strict-sign conclusions. Weak comparator calibration limits the Amazon evidence to fitted-surface alignment rather than latent-preference recovery.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.LG, nuovi lavori del giorno).
Categorie arXiv: Machine Learning (cs.LG) ; Information Retrieval (cs.IR).
Pagina del lavoro: https://arxiv.org/abs/2611.04997 — 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 2026-11-04T21:38:49.742Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).