Metraipolis › archive › dc_0bf102b770c0

Statistical Convergence of Transformer Encoder-Accelerated Robust Reinforcement Learning — arXiv:2609.23775

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.23775
License (as archived)
CC0-1.0
Archived
2026-10-09T07:07:20.162Z
SHA-256 of the archived text
787fbb8f54af5f91aa0621baa2d9bf6959d477008e959d5674fd2b2cc1485199
Stable URL of this page
https://metraipolis.com/documento/dc_0bf102b770c0

How to cite

Statistical Convergence of Transformer Encoder-Accelerated Robust Reinforcement Learning — arXiv:2609.23775 — arXiv (en), https://arxiv.org/abs/2609.23775 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_0bf102b770c0 (Metraipolis archive dc_0bf102b770c0, sha256 787fbb8f54af5f91aa0621baa2d9bf6959d477008e959d5674fd2b2cc1485199)
BibTeX
@misc{dc0bf102b770c0,
  title = {Statistical Convergence of Transformer Encoder-Accelerated Robust Reinforcement Learning — arXiv:2609.23775},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.23775},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_0bf102b770c0; sha256 787fbb8f54af5f91aa0621baa2d9bf6959d477008e959d5674fd2b2cc1485199},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.23775. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Obtaining the optimal action-value function in Markov decision processes is computationally intensive in large state--action spaces. In this study, we present statistically rigorous convergence results for a robust reinforcement learning algorithm warm-started by a transformer-based action-value function prediction, where natural language prompts encode task specifications. Our framework adopts the R-contamination model to characterize uncertainty in the state transition kernel, and employs conformal prediction to certify convergence via trajectory-level nonconformity scores constructed from the contracting Bellman residual. The resulting conformal quantile bounds the gap between the running and optimal action-value functions simultaneously over all iterations, thereby yielding a pre-certified stopping rule that requires little knowledge of the true transition kernel. Numerical case studies on perturbed maze environments of varying size and contamination level confirm that the transformer-based warm start measurably reduces the initial error and accelerates convergence, while the proposed conformal bounds track the true error trajectory more tightly than existing guarantees.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-22 in cs.LG).
Categorie arXiv: Machine Learning (cs.LG) ; Systems and Control (eess.SY).
Pagina del lavoro: https://arxiv.org/abs/2609.23775 — 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 2026-10-09T06:58:09.326Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-09-22?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).