Metraipolis › archive › dc_6ddf705ebbde

Momentum-Coupled Rubric Adaptation for Detailed Image Captioning — arXiv:2609.36893

Source: arXiv (en)
Original: https://arxiv.org/abs/2609.36893
License (as archived): CC0-1.0
Archived: 2026-10-07T16:56:43.441Z
SHA-256 of the archived text: aea95cb6589333cab8820af632c5c4fb7e8b93eaea15883f77438e49fd2d0a58
Stable URL of this page: https://metraipolis.com/documento/dc_6ddf705ebbde

How to cite

Momentum-Coupled Rubric Adaptation for Detailed Image Captioning — arXiv:2609.36893 — arXiv (en), https://arxiv.org/abs/2609.36893 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_6ddf705ebbde (Metraipolis archive dc_6ddf705ebbde, sha256 aea95cb6589333cab8820af632c5c4fb7e8b93eaea15883f77438e49fd2d0a58)

BibTeX

@misc{dc6ddf705ebbde,
  title = {Momentum-Coupled Rubric Adaptation for Detailed Image Captioning — arXiv:2609.36893},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.36893},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_6ddf705ebbde; sha256 aea95cb6589333cab8820af632c5c4fb7e8b93eaea15883f77438e49fd2d0a58},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.36893. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Detailed image captioning requires accurate and comprehensive descriptions of fine-grained visual content, yet caption quality spans factual accuracy, information coverage, and clarity. Compared with conventional methods that rely mainly on high-quality supervision or holistic rewards, rubric-based reinforcement learning decomposes these requirements into explicit criteria and provides targeted, structured feedback. However, existing methods often use separate models for caption generation, rubric construction, and judging, which may lead to inconsistent interpretations across roles. Some dynamic rubric methods alternate updates between the caption policy and rubric generator while keeping the judge fixed, but staged optimization may still leave rubric construction and judging out of step with policy optimization. We propose MoCo Rubric, a two-stage framework that coordinates these roles. First, role-conditioned, shared-parameter multi-task supervised fine-tuning equips a single vision--language model to serve as the Caption Policy, Rubric Generator, and Rubric Judge. Then, the Generator constructs rubrics online from captions sampled by the current Policy, reference captions, and image evidence. The Judge provides rubric-based rewards, and only the Policy receives GRPO updates. As Policy updates change the candidates being evaluated, we use an exponential moving average of the Policy parameters to update one momentum model shared by the Generator and Judge. This gradual transfer lets both rubric roles track Policy updates without separate RL optimization while smoothing parameter changes that could disrupt their rubric capabilities under direct synchronization. Across five captioning benchmarks, MoCo Rubric achieves an average pairwise win rate of 72.83\%, the best mean rank in blind ranking, and the highest average score in caption-based question answering.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-30 in cs.CL).
Categorie arXiv: Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2609.36893 — 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 2026-10-07T16:27:03.245Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-09-30?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).