Metraipolis › archive › dc_33f3e36fc59f

Dual-Stream Cross-Anchor Correction Grounding Long-Form Captions and the Domain Limits of Object-Level Anchors — arXiv:2608.12746

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.12746
License (as archived)
CC0-1.0
Archived
2026-10-19T04:24:41.180Z
SHA-256 of the archived text
924674373d0fb7a8c1a83bee83e5d850d810b34d89ab08827a255c39f735d01e
Stable URL of this page
https://metraipolis.com/documento/dc_33f3e36fc59f

How to cite

Dual-Stream Cross-Anchor Correction Grounding Long-Form Captions and the Domain Limits of Object-Level Anchors — arXiv:2608.12746 — arXiv (en), https://arxiv.org/abs/2608.12746 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_33f3e36fc59f (Metraipolis archive dc_33f3e36fc59f, sha256 924674373d0fb7a8c1a83bee83e5d850d810b34d89ab08827a255c39f735d01e)
BibTeX
@misc{dc33f3e36fc59f,
  title = {Dual-Stream Cross-Anchor Correction Grounding Long-Form Captions and the Domain Limits of Object-Level Anchors — arXiv:2608.12746},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.12746},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_33f3e36fc59f; sha256 924674373d0fb7a8c1a83bee83e5d850d810b34d89ab08827a255c39f735d01e},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.12746. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Object hallucination in multimodal large language models arises when language priors and corpus co-occurrence bias outweigh the visual evidence, with nothing tying an object mention to the image. Most remedies intervene at decoding time, yet under a unified protocol their benefit is confined to short captions; supervised fine-tuning (SFT) on a detail-rich corpus lengthens captions, but over forty percent still name absent objects. This paper proposes Dual-Stream Cross-Anchor Correction (DSCC). Unlike work that post-processes decoding, DSCC injects object-level visual anchors into the language model itself during fine-tuning: a perception stream aligns object-level hidden states at an intermediate layer to frozen text anchors by a bidirectional contrastive objective; a cognition stream lets deeper layers query those anchors by cross-attention at every generation step; and a two-stage curriculum gate couples them, making evidence retrieval a structural constraint on generation. Under one backbone and one scoring protocol, experiments span long-caption hallucination, object-existence discrimination and cross-domain generalisation, with vanilla SFT on the same corpus and schedule as a length- and density-matched control separating the data effect from the architectural gain. DSCC alone reaches the long-caption, low-hallucination region: captions roughly 1.9 times the baseline length at 88.19% precision per object mention, the highest under a density-independent criterion. Ablations expose a synergy: the perception stream alone degrades precision yet reverses sign when stacked on the cognition stream. No universal superiority is claimed: three out-of-domain benchmarks yield a predictable, falsifiable domain-conditionality, the synergy being bound to the anchors' semantic domain and breaking on charts and illusions.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-14 in cs.CL).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2608.12746 — 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 2026-10-19T04:17:42.409Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-08-14?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).