Metraipolis › archive › dc_ea2eca3e994c

DARE to Mitigate Hallucination: Dual-path Auto-Regressive-aware Editing — arXiv:2609.36440

Source: arXiv (en)
Original: https://arxiv.org/abs/2609.36440
License (as archived): CC0-1.0
Archived: 2026-10-07T18:38:11.034Z
SHA-256 of the archived text: 46b78bf8f6b2c014926db4c6a9b8405426ee0ea835523948a30c141b3b86069c
Stable URL of this page: https://metraipolis.com/documento/dc_ea2eca3e994c

How to cite

DARE to Mitigate Hallucination: Dual-path Auto-Regressive-aware Editing — arXiv:2609.36440 — arXiv (en), https://arxiv.org/abs/2609.36440 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_ea2eca3e994c (Metraipolis archive dc_ea2eca3e994c, sha256 46b78bf8f6b2c014926db4c6a9b8405426ee0ea835523948a30c141b3b86069c)

BibTeX

@misc{dcea2eca3e994c,
  title = {DARE to Mitigate Hallucination: Dual-path Auto-Regressive-aware Editing — arXiv:2609.36440},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.36440},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_ea2eca3e994c; sha256 46b78bf8f6b2c014926db4c6a9b8405426ee0ea835523948a30c141b3b86069c},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.36440. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Large vision-language models (LVLMs) have recently achieved remarkable progress across multimodal tasks, yet object hallucination remains a persistent challenge where models generate descriptions inconsistent with the visual input. Recent work mitigates hallucinations through training-free representation editing, typically by constructing hallucination-related directions from teacher-forcing (TF) contrasts between hallucinated and truthful responses. However, LVLMs operate through autoregressive (AR) decoding during generation, raising the question of whether TF-based analysis fully reflects the generation dynamics that lead to hallucinated outputs. In this paper, we analyze the relationship between TF-based editing and AR generation behavior and find that TF-based editing alone may be insufficient to capture both decoding dynamics and multimodal interactions associated with hallucinations. To address this limitation, we propose DARE (Dual-path Auto-Regressive-aware Editing), a hybrid hallucination editing framework that integrates two complementary contrast pathways: textual contrasts and image contrasts, together with autoregressive-aware representation signals. Specifically, DARE constructs hallucination editing directions from (1) TF-based textual contrasts, (2) AR-aware representation transitions during decoding, and (3) controlled visual differences between paired images. Extensive experiments on multiple LVLM hallucination benchmarks demonstrate that DARE consistently reduces object hallucinations while preserving multimodal perception capability and inference efficiency. Our implementation code is available at this https URL .

Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-09-30 in cs.CV).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2609.36440 — 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 2026-10-07T18:04:24.704Z dalla pagina https://arxiv.org/catchup/cs.CV/2026-09-30?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).