Metraipolis › archive › dc_095fcafbc443

Task Inference Beyond Least Squares in Behavioral Foundation Models — arXiv:2610.05350

Source: arXiv
Original: https://arxiv.org/abs/2610.05350
License (as archived): CC0-1.0
Archived: 2026-10-06T05:31:04.271Z
SHA-256 of the archived text: 868ea35b87b86f4c2a91c816a638c97562d647fe70f26d33a7359aa5e542d00e
Stable URL of this page: https://metraipolis.com/documento/dc_095fcafbc443

How to cite

Task Inference Beyond Least Squares in Behavioral Foundation Models — arXiv:2610.05350 — arXiv, https://arxiv.org/abs/2610.05350 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_095fcafbc443 (Metraipolis archive dc_095fcafbc443, sha256 868ea35b87b86f4c2a91c816a638c97562d647fe70f26d33a7359aa5e542d00e)

BibTeX

@misc{dc095fcafbc443,
  title = {Task Inference Beyond Least Squares in Behavioral Foundation Models — arXiv:2610.05350},
  howpublished = {arXiv},
  url = {https://arxiv.org/abs/2610.05350},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_095fcafbc443; sha256 868ea35b87b86f4c2a91c816a638c97562d647fe70f26d33a7359aa5e542d00e},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv, https://arxiv.org/abs/2610.05350. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Behavioral Foundation Models (BFMs) aim to solve a wide range of downstream tasks without test-time policy learning by inferring a task vector from the reward function. While efficient, the retrieved policies are often suboptimal because of how this task vector is inferred, typically with ordinary least squares (OLS). OLS minimizes reward reconstruction error but leaves the ordering of rewards unconstrained, which can bias the successor measure of the retrieved zero-shot policy away from that of the optimal policy. In this work, we propose BLS, an efficient test-time inference method that balances minimizing reward reconstruction error with reducing successor-measure mismatch. Theoretically, we provide a suboptimality gap upper bound characterized by both successor-measure and reward-function residuals. Empirically, we evaluate BLS on top of state-of-the-art BFMs across benchmarks for locomotion, manipulation, and humanoid control. BLS outperforms existing task inference baselines with negligible computational overhead. Project page: this https URL

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/2610.05350 — 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 2026-10-06T05:27:05.789Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).