Metraipolis › archive › dc_e752f387cc79
LEGO-OPD: Factorized Teacher Composition for Multimodal On-Policy Distillation — arXiv:2610.00333
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.00333
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-06T12:08:45.681Z
- SHA-256 of the archived text
535eb36d843ed3d62dc9d5d70e5f11a3b49b14b72fb2c3dd550af01c355642b4- Stable URL of this page
https://metraipolis.com/documento/dc_e752f387cc79
How to cite
LEGO-OPD: Factorized Teacher Composition for Multimodal On-Policy Distillation — arXiv:2610.00333 — arXiv (en), https://arxiv.org/abs/2610.00333 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_e752f387cc79 (Metraipolis archive dc_e752f387cc79, sha256 535eb36d843ed3d62dc9d5d70e5f11a3b49b14b72fb2c3dd550af01c355642b4)
BibTeX
@misc{dce752f387cc79,
title = {LEGO-OPD: Factorized Teacher Composition for Multimodal On-Policy Distillation — arXiv:2610.00333},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.00333},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_e752f387cc79; sha256 535eb36d843ed3d62dc9d5d70e5f11a3b49b14b72fb2c3dd550af01c355642b4},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.00333. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Multimodal on-policy distillation (OPD) aims to improve visual grounding while preserving the strong reasoning capabilities of language models. Recent multi-teacher approaches combine LLM and VLM teachers to provide complementary supervision. However, directly using a VLM's full predictive distribution entangles its visual grounding signal with its own language prior, preventing the grounding information from being transferred independently. Conversely, increasing the strength of visual supervision can improve perception but may overemphasize visual evidence and degrade language reasoning. To address this trade-off, we introduce LEGO-OPD, which selectively composes factors from a Language Expert and a Grounding expert into One teacher distribution for multimodal OPD. Under a generalized Bayesian formulation, the language expert provides a prior over candidate tokens, while the grounding expert contributes a visual likelihood that updates this prior, rather than transferring its complete predictive distribution. This factorized composition allows language reasoning and visual grounding to be controlled independently. We further introduce adaptive calibration to determine how strongly the visual likelihood should update the language prior at each decoding prefix. Specifically, LEGO-OPD uses the grounding expert's image-induced prediction shift as a prefix-dependent reference, preventing both insufficient and excessive visual supervision. Experiments with Qwen3 models show that LEGO-OPD consistently outperforms the evaluated single- and multi-teacher OPD baselines on both multimodal and text-only reasoning tasks. Moreover, it improves the initial student's visual perception while preserving text-only reasoning. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-10-02 in cs.AI). Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Machine Learning (cs.LG). Pagina del lavoro: https://arxiv.org/abs/2610.00333 — 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 2026-10-06T10:21:28.048Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-10-02?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).