Metraipolis › archive › dc_a22bc36f6240
Selective Agent Guidance via Entropy: Learning Autonomous Policies from Imperfect VLM Teachers — arXiv:2609.01567
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.01567
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-17T06:32:31.716Z
- SHA-256 of the archived text
bde1cabaab4fd66f941262013213da32af8889d1e4330d63948c34b06ff3bfcb- Stable URL of this page
https://metraipolis.com/documento/dc_a22bc36f6240
How to cite
Selective Agent Guidance via Entropy: Learning Autonomous Policies from Imperfect VLM Teachers — arXiv:2609.01567 — arXiv (en), https://arxiv.org/abs/2609.01567 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_a22bc36f6240 (Metraipolis archive dc_a22bc36f6240, sha256 bde1cabaab4fd66f941262013213da32af8889d1e4330d63948c34b06ff3bfcb)
BibTeX
@misc{dca22bc36f6240,
title = {Selective Agent Guidance via Entropy: Learning Autonomous Policies from Imperfect VLM Teachers — arXiv:2609.01567},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.01567},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_a22bc36f6240; sha256 bde1cabaab4fd66f941262013213da32af8889d1e4330d63948c34b06ff3bfcb},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.01567. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Vision-Language Models (VLMs) provide useful priors for interactive decision-making, but using them directly as policies is expensive and brittle: they must be queried at every step, do not improve from environment interaction, and can repeat systematic errors. We study how to learn a cheap autonomous policy from an online, expensive, and imperfect but informative VLM teacher. We propose SAGE (Selective Agent Guidance via Entropy), a framework that queries a VLM only when the learner is uncertain, executes the suggested action during training, and distills guidance into a lightweight Reinforcement Learning (RL) policy. Because VLM advice is not always reliable, SAGE can weight teacher-action distillation using environment-derived advantages rather than treating all suggestions as equally useful. Across sparse-reward visual reasoning and navigation tasks, SAGE learns policies that act without VLM guidance at evaluation time and improves over unguided RL in several environments, including settings where the learned policy exceeds its VLM teacher. The results show that selective guidance is most beneficial when the VLM can help the agent discover high-reward trajectories, and less useful when unguided exploration already succeeds or teacher actions do not lead to informative experience. SAGE also reduces VLM usage by prompting the teacher only on a fraction of training steps and requiring no VLM calls at deployment. Overall, our results suggest that VLMs don't need to be used as fixed policies to be useful; they can instead act as temporary, imperfect sources of guidance whose value is tested and internalized through interaction. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-02 in cs.AI). Categorie arXiv: Artificial Intelligence (cs.AI) ; Computation and Language (cs.CL); Machine Learning (cs.LG). Pagina del lavoro: https://arxiv.org/abs/2609.01567 — 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 2026-10-17T06:15:12.703Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-02?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).