Metraipolis › archive › dc_ba3147470bbb

Who Leads Now? Token-Level Modality Arbitration for Chart-to-Code Generation — arXiv:2608.15510

Source: arXiv (en)
Original: https://arxiv.org/abs/2608.15510
License (as archived): CC0-1.0
Archived: 2026-10-18T17:25:00.009Z
SHA-256 of the archived text: 327380360818d5df04394e939621152d35c3b5ed4767f47f561e9accab1ae783
Stable URL of this page: https://metraipolis.com/documento/dc_ba3147470bbb

How to cite

Who Leads Now? Token-Level Modality Arbitration for Chart-to-Code Generation — arXiv:2608.15510 — arXiv (en), https://arxiv.org/abs/2608.15510 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_ba3147470bbb (Metraipolis archive dc_ba3147470bbb, sha256 327380360818d5df04394e939621152d35c3b5ed4767f47f561e9accab1ae783)

BibTeX

@misc{dcba3147470bbb,
  title = {Who Leads Now? Token-Level Modality Arbitration for Chart-to-Code Generation — arXiv:2608.15510},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.15510},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_ba3147470bbb; sha256 327380360818d5df04394e939621152d35c3b5ed4767f47f561e9accab1ae783},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.15510. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Chart-to-code generation requires a model to read the fine-grained visual details of a chart and write executable code that reproduces it. Existing chart-to-code methods either train visual and coding abilities separately, or fine-tune on chart-to-code data with the two abilities entangled. Neither strategy accounts for the distinct nature of the two abilities or the interference that arises when they are optimized together. We propose MoCA (Mixture of Cross-modal Arbitration), which separates the two abilities rather than blending them. MoCA is built on Cross-modal Arbitration Block (CAB), which maintains a visual branch and a code branch as two distinct pathways, and a lightweight arbiter that arbitrates their relative contributions at every layer and generated token. We train MoCA in two stages: a supervised warm-up on self-distilled reasoning trajectories that decomposes visual understanding into explicit steps, followed by reinforcement learning with rewards on both the reasoning process and the final code. Analysis shows that the arbiter learns structured rather than arbitrary allocations, with expert contributions varying systematically across tokens, layers, and instances. Across three benchmarks, MoCA delivers competitive performance against general-domain and chart-specialized models. Ablation results show that the gains cannot be attributed to a larger model size alone, but instead arise from the joint contributions of complementary visual and code branch initialization and input-conditioned arbitration through CAB.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-18 in cs.AI).
Categorie arXiv: Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2608.15510 — 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 2026-10-18T16:29:13.807Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-08-18?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).