Gated Relational Alignment via Confidence-based Distillation for Efficient VLMs 文章

ArXiv CS.CV2026-05-26NEWSen作者: Yanlong Chen, Amirhossein Habibian, Luca Benini, Yawei Li
人工智能
查看原文 →
关系图谱

Gated Relational Alignment via Confidence-based Distillation for Efficient VLMs · 相关技术

相关技术

ORMODEPLAlanguage model研究法辛烷值多模态后训练alignmentUCTTokenStraight-Through EstimatorQuantization-Aware TrainingQATParts-of-Speech (POS) tagsOWLVision-Language ModelsVLMKnowledge DistillationKernelHISForFFIEffort Metric AttentionDAPTANN