Metraipolis › archive › dc_85d3d3e11568
Reinforcement Learning via Advantage Guidance — arXiv:2610.19783
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.19783
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-15T14:17:47.449Z
- SHA-256 of the archived text
c16522d4bdb2a029fb7dcd0a77955e7ccf85eb9073caf53a077f75460f8049cb- Stable URL of this page
https://metraipolis.com/documento/dc_85d3d3e11568
How to cite
Reinforcement Learning via Advantage Guidance — arXiv:2610.19783 — arXiv (en), https://arxiv.org/abs/2610.19783 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_85d3d3e11568 (Metraipolis archive dc_85d3d3e11568, sha256 c16522d4bdb2a029fb7dcd0a77955e7ccf85eb9073caf53a077f75460f8049cb)
BibTeX
@misc{dc85d3d3e11568,
title = {Reinforcement Learning via Advantage Guidance — arXiv:2610.19783},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.19783},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_85d3d3e11568; sha256 c16522d4bdb2a029fb7dcd0a77955e7ccf85eb9073caf53a077f75460f8049cb},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.19783. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Reinforcement learning (RL) holds the promise of learning a policy that can outperform what is shown in the data. However, in practice, such systems suffer from instability and hyperparameter sensitivity. In contrast, supervised learning provides a stable and scalable training objective, but is fundamentally limited by the quality of the demonstrated behavior. In this paper, we ask whether we can perform policy optimization with something as simple and stable as supervised learning but as performant as RL. Unfortunately, naively training the policy with supervised learning and conditioning on future rewards or advantages typically underperform dedicated RL methods. We introduce a new policy extraction method, Advantage Guidance, that retains the simple supervised learning structure with advantage conditioning while outperforming prior RL methods through effective conditional training and contrastive diffusion guidance during sampling. We show that Advantage Guidance can significantly improve a model's ability to respond to conditional advantage signals, and the contrastive guidance during sampling can use the learned advantage-conditioned policy to target the KL-regularized optimal policy. Empirically, we find that Advantage Guidance significantly outperforms prior state-of-the-art RL methods on 26 long-horizon, sparse reward tasks across datasets with varying degrees of optimality. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.LG, nuovi lavori del giorno). Categorie arXiv: Machine Learning (cs.LG). Pagina del lavoro: https://arxiv.org/abs/2610.19783 — 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 2026-10-15T13:53:35.643Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).