paper-with-me

홈 › Papers

MRFD: Multi-Region Fusion Decoding with Self-Consistency for Mitigating Hallucinations in LVLMs

2025-08-14 · Haonan Ge, Yiwei Wang, Ming-Hsuan Yang, Yujun Cai arxiv

Large Vision-Language Models (LVLMs) have shown strong performance across multimodal tasks. However, they often produce hallucinations -- text that is inconsistent with visual input, due to the limited ability to verify information in different regions of the image. To address this, we propose Multi-Region Fusion Decoding (MRFD), a training-free decoding method that improves factual grounding by modeling inter-region consistency. MRFD identifies salient regions using cross-attention, generates initial responses for each, and computes reliability weights based on Jensen-Shannon Divergence (JSD) among the responses. These weights guide a consistency-aware fusion of per-region predictions, using region-aware prompts inspired by Chain-of-Thought reasoning. Experiments across multiple LVLMs and benchmarks show that MRFD significantly reduces hallucinations and improves response factuality without requiring model updates.

📄 PDF Abstract BibTeX arXiv:2508.10264

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

Consistent Multiple Sequence Decoding

2020-04-02 · Bicheng Xu, Leonid Sigal

Sequence decoding is one of the core components of most visual-lingual models. However, typical neural decoders when faced with decoding multiple, possibly correlated, sequences of tokens resort to simple independent dec…

DecoderDiversityGraph Neural NetworkImage Captioning

BaRISTA: Brain Scale Informed Spatiotemporal Representation of Human Intracranial Neural Activity

2025-12-13 · Lucine L. Oganesian, Saba Hashemi, Maryam M. Shanechi arxiv

Intracranial recordings have opened a unique opportunity to simultaneously measure activity across multiregional networks in the human brain. Recent works have focused on developing transformer-based neurofoundation mode…

Self Speculative Decoding for Diffusion Large Language Models

2025-10-05 · Yifeng Gao, Ziang Ji, Yuxuan Wang, Biqing Qi 외 arxiv

Diffusion-based Large Language Models (dLLMs) have emerged as a competitive alternative to autoregressive models, offering unique advantages through bidirectional attention and parallel generation paradigms. However, the…

GeoBlock: Inferring Block Granularity from Dependency Geometry in Diffusion Language Models

2026-03-04 · Lipeng Wan, Junjie Ma, Jianhui Gu, Zeyang Liu 외 arxiv

Block diffusion enables efficient parallel refinement in diffusion language models, but its decoding behavior depends critically on block size. Existing block-sizing strategies rely on fixed rules or heuristic signals an…

S2D2: Fast Decoding for Diffusion LLMs via Training-Free Self-Speculation

2026-03-26 · Ligong Han, Hao Wang, Han Gao, Kai Xu 외 arxiv

Block-diffusion language models offer a promising path toward faster-than-autoregressive generation by combining block-wise autoregressive decoding with within-block parallel denoising. However, in the few-step regime ne…