Zhichun Wang

Xiaoju Ye, Zhichun Wang, Jingyuan Wang

Limited by the context window size of Large Language Models(LLMs), handling various tasks with input tokens exceeding the upper limit has been challenging, whether it is a simple direct retrieval task or a complex multi-hop reasoning task. Although various methods have been proposed to enhance the long-context processing capabilities of LLMs, they either incur substantial post-training costs, or require additional tool modules(e.g.,RAG), or have not shown significant improvement in realistic tasks. Our work observes the correlation between the attention distribution and generated answers across each layer, and establishes the attention allocation aligns with retrieval-augmented capabilities through experiments. Drawing on the above insights, we propose a novel method InfiniRetri that leverages the LLMs's own attention information to enable accurate retrieval across inputs of infinitely length. Our evaluations indicate that InfiniRetri achieves 100% accuracy in the Needle-In-a-Haystack(NIH) test over 1M tokens using a 0.5B parameter model, surpassing other method or larger models and setting a new state-of-the-art(SOTA). Moreover, our method achieves significant performance improvements on real-world benchmarks, with a maximum 288% improvement. In addition, InfiniRetri can be applied to any Transformer-based LLMs without additional training and substantially reduces inference latency and compute overhead in long texts. In summary, our comprehensive studies show InfiniRetri's potential for practical applications and creates a paradigm for retrievaling information using LLMs own capabilities under infinite-length tokens. Code will be released in link.

Xianpei Han, Zhichun Wang, Jiangtao Zhang et al.

Knowledge graph models world knowledge as concepts, entities, and the relationships between them, which has been widely used in many real-world tasks. CCKS 2019 held an evaluation track with 6 tasks and attracted more than 1,600 teams. In this paper, we give an overview of the knowledge graph evaluation tract at CCKS 2019. By reviewing the task definition, successful methods, useful resources, good strategies and research challenges associated with each task in CCKS 2019, this paper can provide a helpful reference for developing knowledge graph applications and conducting future knowledge graph researches.

Zhichun Wang, Juanzi Li

Recently, several large-scale RDF knowledge bases have been built and applied in many knowledge-based applications. To further increase the number of facts in RDF knowledge bases, logic rules can be used to predict new facts based on the existing ones. Therefore, how to automatically learn reliable rules from large-scale knowledge bases becomes increasingly important. In this paper, we propose a novel rule learning approach named RDF2Rules for RDF knowledge bases. RDF2Rules first mines frequent predicate cycles (FPCs), a kind of interesting frequent patterns in knowledge bases, and then generates rules from the mined FPCs. Because each FPC can produce multiple rules, and effective pruning strategy is used in the process of mining FPCs, RDF2Rules works very efficiently. Another advantage of RDF2Rules is that it uses the entity type information when generates and evaluates rules, which makes the learned rules more accurate. Experiments show that our approach outperforms the compared approach in terms of both efficiency and accuracy.