Metraipolis › archive › dc_dd093ce87b09
Transformer-Based Reinforcement Learning for Phase I Dose Allocation — arXiv:2611.00395
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2611.00395
- License (as archived)
- CC0-1.0
- Archived
- 2026-11-03T18:39:23.258Z
- SHA-256 of the archived text
46636fb1340f8eb37a5463f584b683683d48807451ffeb374c95cd6f2d78ede0- Stable URL of this page
https://metraipolis.com/documento/dc_dd093ce87b09
How to cite
Transformer-Based Reinforcement Learning for Phase I Dose Allocation — arXiv:2611.00395 — arXiv (en), https://arxiv.org/abs/2611.00395 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_dd093ce87b09 (Metraipolis archive dc_dd093ce87b09, sha256 46636fb1340f8eb37a5463f584b683683d48807451ffeb374c95cd6f2d78ede0)
BibTeX
@misc{dcdd093ce87b09,
title = {Transformer-Based Reinforcement Learning for Phase I Dose Allocation — arXiv:2611.00395},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2611.00395},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_dd093ce87b09; sha256 46636fb1340f8eb37a5463f584b683683d48807451ffeb374c95cd6f2d78ede0},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.00395. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Phase I dose-finding trials must learn from sparse toxicity observations while limiting exposure to overly toxic doses. We develop a single reinforcement-learning policy for trials with different numbers of ordered dose levels. Each dose is represented as a token containing its normalized position, accumulated patient count, observed dose-limiting toxicities (DLTs), and current-dose indicator. A Transformer encoder summarizes the dose tokens, trial-conditioned cross-attention produces a fixed-dimensional trial representation, and a three-action policy chooses de-escalation, staying, or escalation. The final maximum tolerated dose (MTD) decision is deliberately kept outside the learned policy: after the planned sample size, weighted isotonic regression and a prespecified safety cap select the dose closest to the target toxicity probability. In the simulations, toxicity-penalized policies reduce DLT burden while unpenalized policies improve selection accuracy at the cost of greater toxicity. These results demonstrate a controllable safety--accuracy trade-off but do not establish superiority over conventional dose-finding designs. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.LG, nuovi lavori del giorno). Categorie arXiv: Machine Learning (stat.ML) ; Machine Learning (cs.LG); Applications (stat.AP). Pagina del lavoro: https://arxiv.org/abs/2611.00395 — 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 2026-11-03T18:12:54.687Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).