Generating Event-oriented Attribution for Movies via Two-Stage Prefix-Enhanced Multimodal LLM

no code implementations • 14 Sep 2024 • Yuanjie Lyu, Tong Xu, Zihan Niu, Bo Peng, Jing Ke, Enhong Chen

Correspondingly, in the global stage, we strengthen the connections between associated events using an inferential knowledge graph, and design an event-aware prefix that directs the model to focus on associated events rather than all preceding clips, resulting in accurate event attribution.

Retrieve-Plan-Generation: An Iterative Planning and Answering Framework for Knowledge-Intensive LLM Generation

1 code implementation • 21 Jun 2024 • Yuanjie Lyu, Zihan Niu, Zheyong Xie, Chao Zhang, Tong Xu, Yang Wang, Enhong Chen

Despite the significant progress of large language models (LLMs) in various tasks, they often produce factual errors due to their limited internal knowledge.

Answer Generation RAG

CRUD-RAG: A Comprehensive Chinese Benchmark for Retrieval-Augmented Generation of Large Language Models

1 code implementation • 30 Jan 2024 • Yuanjie Lyu, Zhiyu Li, Simin Niu, Feiyu Xiong, Bo Tang, Wenjin Wang, Hao Wu, Huanyong Liu, Tong Xu, Enhong Chen

For each of these CRUD categories, we have developed comprehensive datasets to evaluate the performance of RAG systems.

Knowledge Base Construction Question Answering +2

Faithful Abstractive Summarization via Fact-aware Consistency-constrained Transformer

1 code implementation • CIKM 2022 • Yuanjie Lyu, Chen Zhu, Tong Xu, Zikai Yin, Enhong Chen

To deal with this challenge, in this paper, we propose a novel fact-aware abstractive summarization model, named Entity-Relation Pointer Generator Network (ERPGN).

Abstractive Text Summarization Text Generation