Metraipolis › archive › dc_97070b199ce7

Lookahead Prediction-Guided Adaptive Visual Refocusing for Interleaved Multimodal Reasoning — arXiv:2611.02780

Source
arXiv (en)
Original
https://arxiv.org/abs/2611.02780
License (as archived)
CC0-1.0
Archived
2026-11-03T11:42:38.076Z
SHA-256 of the archived text
4de951fb6fd98a92fec8490244c8518d5b8ed81becddb7576d964b197ee33f3a
Stable URL of this page
https://metraipolis.com/documento/dc_97070b199ce7

How to cite

Lookahead Prediction-Guided Adaptive Visual Refocusing for Interleaved Multimodal Reasoning — arXiv:2611.02780 — arXiv (en), https://arxiv.org/abs/2611.02780 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_97070b199ce7 (Metraipolis archive dc_97070b199ce7, sha256 4de951fb6fd98a92fec8490244c8518d5b8ed81becddb7576d964b197ee33f3a)
BibTeX
@misc{dc97070b199ce7,
  title = {Lookahead Prediction-Guided Adaptive Visual Refocusing for Interleaved Multimodal Reasoning — arXiv:2611.02780},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.02780},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_97070b199ce7; sha256 4de951fb6fd98a92fec8490244c8518d5b8ed81becddb7576d964b197ee33f3a},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.02780. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Interleaved multimodal chain-of-thought reasoning improves the visual reasoning capabilities of multimodal large language models by dynamically incorporating visual evidence during reasoning. However, existing methods primarily select evidence from historical context. When the reasoning focus shifts, history-guided evidence may remain focused on prior regions and fail to cover the visual information required by the next step. We refer to this lag between visual evidence and subsequent reasoning demands as Temporal Evidence Myopia. To address this issue, we propose Lookahead Prediction-guided Adaptive Visual Refocusing (LPAVR), a training-free test-time framework. LPAVR transforms Multi-Token Prediction (MTP) from a decoding acceleration mechanism into a lookahead signal for anticipating upcoming reasoning intent. During MTP decoding, the draft head generates multiple future tokens, while the target model verifies them in parallel and retains the longest prefix consistent with its decoding trajectory. This target-verified prefix contains near-future semantics, which LPAVR uses to construct lookahead-guided visual evidence and assess whether history-guided evidence covers the indicated regions. If the coverage is sufficient, the history-guided evidence is retained; otherwise, LPAVR switches to the lookahead-guided evidence. This better aligns the visual context with upcoming reasoning and improves consistency between draft predictions and target-model decoding. Experiments on multiple multimodal reasoning benchmarks show that LPAVR improves both reasoning accuracy and the acceptance rates of MTP speculative tokens, jointly enhancing visual reasoning and speculative decoding efficiency.

Tema degli esploratori: statistica e scienza dei dati (elenco arXiv cs.CV, nuovi lavori del giorno).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2611.02780 — 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 2026-11-03T11:06:37.590Z dalla pagina https://arxiv.org/list/cs.CV/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).