Metraipolis › archive › dc_5e76ab025c44

Does reflective LLM feedback survive a wet-lab-scale label budget? — arXiv:2611.01908

Source
arXiv (en)
Original
https://arxiv.org/abs/2611.01908
License (as archived)
CC0-1.0
Archived
2026-11-03T10:52:44.222Z
SHA-256 of the archived text
961991b2df9a2bec282d4216267654aaa296e55cd11eb468f3e1fc2dc99fc42e
Stable URL of this page
https://metraipolis.com/documento/dc_5e76ab025c44

How to cite

Does reflective LLM feedback survive a wet-lab-scale label budget? — arXiv:2611.01908 — arXiv (en), https://arxiv.org/abs/2611.01908 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_5e76ab025c44 (Metraipolis archive dc_5e76ab025c44, sha256 961991b2df9a2bec282d4216267654aaa296e55cd11eb468f3e1fc2dc99fc42e)
BibTeX
@misc{dc5e76ab025c44,
  title = {Does reflective LLM feedback survive a wet-lab-scale label budget? — arXiv:2611.01908},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.01908},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_5e76ab025c44; sha256 961991b2df9a2bec282d4216267654aaa296e55cd11eb468f3e1fc2dc99fc42e},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.01908. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Can reflective language-model feedback improve a biological predictor when each cycle buys only one plate of labels? We separate three possible sources of benefit: proposing model edits, choosing measurements, and revising decisions through repeated feedback. We replay $36$ ProteinGym assays, starting with $96$ labels and acquiring $96$ per round for six rounds, with a sealed test set and a shared edit schema. The pre-specified comparison of full LLM control against random edits detects no advantage in test rank correlation ($\Delta z=-0.008$, $95\%$ CI $[-0.034,+0.018]$); equivalence is also unestablished. Exploratory ablations nevertheless identify useful model editing. With random acquisition fixed, LLM edits improve Fisher-transformed test Spearman by $0.043$ over no edits; with acquisition fixed to upper confidence bound, they improve normalized top-variant recovery by $0.023$. LLM acquisition alone does not detectably outperform fixed adaptive rules, and repeated feedback does not detectably outperform one initial recommendation under full control. Thus, the observed gains support constrained specialist editing, while the incremental value of recursive reflection remains unestablished at this label budget.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2611.01908 — 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 2026-11-03T10:49:38.319Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).