Výsledky vyhledávání

Report

BRIEF: Bridging Retrieval and Inference for Multi-hop Reasoning via Compression

Autor: Li, Yuankai, Gu, Jia-Chen, Wu, Di, Chang, Kai-Wei, Peng, Nanyun

Retrieval-augmented generation (RAG) can supplement large language models (LLMs) by integrating external knowledge. However, as the number of retrieved documents increases, the input length to LLMs grows linearly, causing a dramatic increase in laten

Externí odkaz: http://arxiv.org/abs/2410.15277

Zobrazit plný text záznamu

Report

Dissecting Dissonance: Benchmarking Large Multimodal Models Against Self-Contradictory Instructions

Autor: Gao, Jin, Gan, Lei, Li, Yuankai, Ye, Yixin, Wang, Dequan

Large multimodal models (LMMs) excel in adhering to human instructions. However, self-contradictory instructions may arise due to the increasing trend of multimodal interaction and context length, which is challenging for language beginners and vulne

Externí odkaz: http://arxiv.org/abs/2408.01091

Zobrazit plný text záznamu

Akademický článek

Research on the Optimization Strategy of the Speed Ratio Control Parameter of Vehicle Swing Pin Chain CVT

Autor: Guo Wei, Li Yuankai, Zhang Wu, Jia Gaoxiang

Publikováno v: Jixie chuandong, Vol 44, Pp 1-8 (2020)

Speed ratio control is one of the key technologies in transmission control field. The fuzzy PID control method is adopted to simulate the automobile cycle working condition by analyzing the structure and working principle of the swing pin chain CVT�

Externí odkaz: https://doaj.org/article/ff6d5cfcad8649c4845d25f5df148b26