Metraipolis › archive › dc_95bc83ca0ca7

Decision Transformer with Task-Aware Representations for Offline Meta-Reinforcement Learning — arXiv:2611.04010

Source: arXiv (en)
Original: https://arxiv.org/abs/2611.04010
License (as archived): CC0-1.0
Archived: 2026-11-04T18:53:55.344Z
SHA-256 of the archived text: 6a1627e42d83f9a80b9ad715d239817ebb5ec674fc779ef1c04527846a154b58
Stable URL of this page: https://metraipolis.com/documento/dc_95bc83ca0ca7

How to cite

Decision Transformer with Task-Aware Representations for Offline Meta-Reinforcement Learning — arXiv:2611.04010 — arXiv (en), https://arxiv.org/abs/2611.04010 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_95bc83ca0ca7 (Metraipolis archive dc_95bc83ca0ca7, sha256 6a1627e42d83f9a80b9ad715d239817ebb5ec674fc779ef1c04527846a154b58)

BibTeX

@misc{dc95bc83ca0ca7,
  title = {Decision Transformer with Task-Aware Representations for Offline Meta-Reinforcement Learning — arXiv:2611.04010},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.04010},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_95bc83ca0ca7; sha256 6a1627e42d83f9a80b9ad715d239817ebb5ec674fc779ef1c04527846a154b58},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.04010. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

To enable effective generalization in offline meta-RL, it is essential to accurately infer the underlying task through latent task representations, which provide a compact and transferable summary of task-specific characteristics for guiding policy generation. In this paper, we propose Contrastive Task-Aware Decision Transformer (CoTA-DT), a DT-based offline meta-RL framework that learns task representations by jointly integrating four complementary components: (i) contrastive learning from offline trajectories, (ii) alignment with natural-language task descriptions, (iii) one-step supervision via a task-conditioned world model, and (iv) long-term supervision via a task-conditioned Q function. The learned task representations are used as conditioning signals for a DT-based policy, enabling task-aware meta-policy learning. To further support generalization in both zero-shot and few-shot settings, we design a language-grounded, self-adaptive strategy that enables effective and reliable adaptation to unseen tasks. Notably, in CoTA-DT, the pretrained LLM serves as a shared backbone for language encoding, world modeling, Q-function estimation, and policy learning. This unified design facilitates parameter sharing, improving both efficiency and generalization. Extensive experiments across diverse offline meta-RL benchmarks demonstrate that CoTA-DT consistently outperforms prior methods, particularly in challenging zero-shot generalization scenarios.

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/2611.04010 — 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 2026-11-04T18:35:37.415Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).