paper-with-me

홈 › Papers

PaMi-VDPO: Mitigating Video Hallucinations by Prompt-Aware Multi-Instance Video Preference Learning

2025-04-08 · Xinpeng Ding, Kui Zhang, Jinahua Han, Lanqing Hong, Hang Xu, Xiaomeng Li

Direct Preference Optimization (DPO) helps reduce hallucinations in Video Multimodal Large Language Models (VLLMs), but its reliance on offline preference data limits adaptability and fails to capture true video-response misalignment. We propose Video Direct Preference Optimization (VDPO), an online preference learning framework that eliminates the need for preference annotation by leveraging video augmentations to generate rejected samples while keeping responses fixed. However, selecting effective augmentations is non-trivial, as some clips may be semantically identical to the original under specific prompts, leading to false rejections and disrupting alignment. To address this, we introduce Prompt-aware Multi-instance Learning VDPO (PaMi-VDPO), which selects augmentations based on prompt context. Instead of a single rejection, we construct a candidate set of augmented clips and apply a close-to-far selection strategy, initially ensuring all clips are semantically relevant while then prioritizing the most prompt-aware distinct clip. This allows the model to better capture meaningful visual differences, mitigating hallucinations, while avoiding false rejections, and improving alignment. PaMi-VDPOseamlessly integrates into existing VLLMs without additional parameters, GPT-4/human supervision. With only 10k SFT data, it improves the base model by 5.3% on VideoHallucer, surpassing GPT-4o, while maintaining stable performance on general video benchmarks.

📄 PDF Abstract BibTeX arXiv:2504.05810

Code (0)

등록된 구현이 없습니다.

Methods 이 논문이 사용한 방법론

SFT Shrink and Fine-Tune, or SFT, is a type of distillation that avoids explicit distillation by copying parameters to a student student model and then fine-tuning.…
BASE 설명 없음
SET Dynamic Sparse Training method where weight mask is updated randomly periodically

Similar Papers 제목 키워드 기반

Vision-Driven Prompt Optimization for Large Language Models in Multimodal Generative Tasks

2025-01-05 · Leo Franklin, Apiradee Boonmee, Kritsada Wongsuwan

Vision generation remains a challenging frontier in artificial intelligence, requiring seamless integration of visual understanding and generative capabilities. In this paper, we propose a novel framework, Vision-Driven …

Image Generation

Mitigating Hallucination in VideoLLMs via Temporal-Aware Activation Engineering

2025-05-19 · JianFeng Cai, Wengang Zhou, Zongmeng Zhang, Jiale Hong 외

Multimodal large language models (MLLMs) have achieved remarkable progress in video understanding.However, hallucination, where the model generates plausible yet incorrect outputs, persists as a significant and under-add…

Hallucination

VIPAMIN: Visual Prompt Initialization via Embedding Selection and Subspace Expansion

2025-10-18 · Jaekyun Park, Hye Won Chung arxiv

In the era of large-scale foundation models, fully fine-tuning pretrained networks for each downstream task is often prohibitively resource-intensive. Prompt tuning offers a lightweight alternative by introducing tunable…

Visual Prompt Tuning

CPR: Mitigating Large Language Model Hallucinations with Curative Prompt Refinement

2025-10-14 · Jung-Woo Shim, Yeong-Joon Ju, Ji-Hoon Park, Seong-Whan Lee arxiv

Recent advancements in large language models (LLMs) highlight their fluency in generating responses to diverse prompts. However, these models sometimes generate plausible yet incorrect ``hallucinated" facts, undermining …

Listening to Patients: A Framework of Detecting and Mitigating Patient Misreport for Medical Dialogue Generation

2024-10-08 · Lang Qin, Yao Zhang, Hongru Liang, Adam Jatowt 외

Medical Dialogue Systems aim to provide automated healthcare support through patient-agent conversations. Previous efforts typically regard patients as ideal users -- one who accurately and consistently reports their hea…

Dialogue GenerationHallucinationMisconceptionsResponse Generation