Metraipolis › archive › dc_d9bcbd975fdb

Tracing the Scientific Discovery Capacity of LLMs to Their Training Data — arXiv:2611.03617

Source: arXiv (en)
Original: https://arxiv.org/abs/2611.03617
License (as archived): CC0-1.0
Archived: 2026-11-03T18:21:34.188Z
SHA-256 of the archived text: 35e8be4cbb42eb673f1a77dc4d9d694b98205bdbd796809211ddeca9eafcb33b
Stable URL of this page: https://metraipolis.com/documento/dc_d9bcbd975fdb

How to cite

Tracing the Scientific Discovery Capacity of LLMs to Their Training Data — arXiv:2611.03617 — arXiv (en), https://arxiv.org/abs/2611.03617 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_d9bcbd975fdb (Metraipolis archive dc_d9bcbd975fdb, sha256 35e8be4cbb42eb673f1a77dc4d9d694b98205bdbd796809211ddeca9eafcb33b)

BibTeX

@misc{dcd9bcbd975fdb,
  title = {Tracing the Scientific Discovery Capacity of LLMs to Their Training Data — arXiv:2611.03617},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.03617},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_d9bcbd975fdb; sha256 35e8be4cbb42eb673f1a77dc4d9d694b98205bdbd796809211ddeca9eafcb33b},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.03617. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Large language models (LLMs) show growing promise for scientific discovery, yet basic questions remain about what enables reliable discovery. How can we trace the scientific discovery capacity of LLMs to their training data? We study this question as one of missing probability mass and develop a theoretical account in which monofacts-the facts observed once during training-set the discovery capacity. In materials science and regulatory genomics, we directly manipulate monofact rates in training sets and find that more monofacts consistently yield more discoveries across two open-weight model families, sampling strategies, and reasoning budgets. Further experiments show the relationship cannot be explained by training-set diversity alone and can transfer across scientific domains when paired with in-domain data. Together, our results identify monofact rate in training data as a key factor governing discoveries in language models.

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/2611.03617 — 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 2026-11-03T18:09:53.766Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).