Metraipolis › archive › dc_8dbdb8226022

Multimodal In-Context Learning with Explicit Rule Induction — arXiv:2610.41754

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.41754
License (as archived)
CC0-1.0
Archived
2026-11-02T06:59:18.674Z
SHA-256 of the archived text
8523235c75973fa5f7756c8c8a9619d28e196a54f3a9d2b665617884be39b038
Stable URL of this page
https://metraipolis.com/documento/dc_8dbdb8226022

How to cite

Multimodal In-Context Learning with Explicit Rule Induction — arXiv:2610.41754 — arXiv (en), https://arxiv.org/abs/2610.41754 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_8dbdb8226022 (Metraipolis archive dc_8dbdb8226022, sha256 8523235c75973fa5f7756c8c8a9619d28e196a54f3a9d2b665617884be39b038)
BibTeX
@misc{dc8dbdb8226022,
  title = {Multimodal In-Context Learning with Explicit Rule Induction — arXiv:2610.41754},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.41754},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_8dbdb8226022; sha256 8523235c75973fa5f7756c8c8a9619d28e196a54f3a9d2b665617884be39b038},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.41754. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

A central challenge in multimodal in-context learning (MICL) arises from tasks that are ill-defined without context and require uncovering the underlying rules from demonstrations. Existing methods predominantly treat MICL as direct input-output mapping, overlooking the importance of explicitly inducing and reasoning over such rules, limiting both performance and interpretability. We propose ERIM, Explicit Rule-Induction for Multimodal In-Context Learning, a unified framework that enhances MICL by explicitly inducing the latent rules governing demonstration samples. Leveraging the reasoning ability of multi-modal large language model (MLLM), ERIM constructs rule-centric MICL datasets, introduces explicit supervision for rule induction, and fine-tunes MLLM to perform explicit rule-induction via a two-stage training strategy that combines mixed cold-start supervised fine-tuning with reinforcement learning. Extensive experiments show consistent performance improvements across diverse MICL settings and demonstrate that correctly inducing underlying rules directly leads to stronger and more stable performance.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2610.41754 — 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 2026-11-02T06:56:58.314Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).