Metraipolis › archive › dc_65dfcbd93b46

Two Uncertainties Are Not One: Disagreement versus Decision-Risk in Cheap Structured Content Moderation — arXiv:2610.33258

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.33258
License (as archived)
CC0-1.0
Archived
2026-10-27T08:11:12.260Z
SHA-256 of the archived text
fbfb5dabc0c5ec503f580aae34e8024fb6acd7e6476ed0faa34e4e550a61bcb4
Stable URL of this page
https://metraipolis.com/documento/dc_65dfcbd93b46

How to cite

Two Uncertainties Are Not One: Disagreement versus Decision-Risk in Cheap Structured Content Moderation — arXiv:2610.33258 — arXiv (en), https://arxiv.org/abs/2610.33258 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_65dfcbd93b46 (Metraipolis archive dc_65dfcbd93b46, sha256 fbfb5dabc0c5ec503f580aae34e8024fb6acd7e6476ed0faa34e4e550a61bcb4)
BibTeX
@misc{dc65dfcbd93b46,
  title = {Two Uncertainties Are Not One: Disagreement versus Decision-Risk in Cheap Structured Content Moderation — arXiv:2610.33258},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.33258},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_65dfcbd93b46; sha256 fbfb5dabc0c5ec503f580aae34e8024fb6acd7e6476ed0faa34e4e550a61bcb4},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.33258. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Content moderation compresses a graded, contested human judgment into a binary act (leave up / take down) over billions of items, at a scale where running a frontier LLM on every post is uneconomical. A new class of frozen "System One" models returns typed, calibrated decisions instead of text at ~USD 0.000025 per call and ~100 ms. We run the first independent audit of one such model on 142,658 human-labeled comments, and benchmark it against 14 frontier reasoning models (the GPT-5.x, Claude, Gemini 3, and Grok families) on a shared probe set. Two findings stand out. First, the cheap model is accuracy-competitive: its zero-shot classification AUROC (0.83) is not statistically separable (paired bootstrap, n=56; a failure to reject parity, not proof of equivalence) from 13 of the 14 frontier systems while costing 1-3 orders of magnitude less, and it is no worse calibrated than several of them. Second, "confidence" conflates two uncertainties that come apart: human disagreement (where annotators split on satire and coded language) and decision risk (where the automatic action is costly-wrong). Under an asymmetric moderation loss these call for opposite escalation signals, so no single confidence number serves both. Because the model prices extra questions at nearly zero, decomposing the judgment into parallel sub-questions and reading their spread recovers a human-disagreement signal that the model's native confidence misses -- but this remedy is specific: it helps the weakest, disagreement-blind models and only when the decomposition aligns with the sources of disagreement; a specialist toxicity model and the strongest LLMs do not need it, and the effect is distribution-sensitive. We will release the instrument, the cached runs, and the analysis.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.CL, nuovi lavori del giorno).
Categorie arXiv: Computation and Language (cs.CL) ; Computational Engineering, Finance, and Science (cs.CE); Logic in Computer Science (cs.LO).
Pagina del lavoro: https://arxiv.org/abs/2610.33258 — 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 2026-10-27T07:56:00.186Z dalla pagina https://arxiv.org/list/cs.CL/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).