Retrieval-augmented generation (RAG) has been extensively employed to mitigate hallucinations in large language models (LLMs). However, existing methods for multi-hop reasoning tasks often lack global planning, increasing the risk of falling into local reasoning impasses. Insufficient exploitation of retrieved content and the neglect of latent clues fail to ensure the accuracy of reasoning outcomes. To overcome these limitations, we propose Recursive Evaluation and Adaptive Planning (REAP), whose core idea is to explicitly maintain structured sub-tasks and facts related to the current task through the Sub-task Planner (SP) and Fact Extractor (FE) modules. SP maintains a global perspective, guiding the overall reasoning direction and evaluating the task state based on the outcomes of FE, enabling dynamic optimization of the task-solving trajectory. FE performs fine-grained analysis over retrieved content to extract reliable answers and clues. These two modules incrementally enrich a logically coherent representation of global knowledge, enhancing the reliability and the traceability of the reasoning process. Furthermore, we propose a unified task paradigm design that enables effective multi-task fine-tuning, significantly enhancing SP's performance on complex, data-scarce tasks. We conduct extensive experiments on multiple public multi-hop datasets, and the results demonstrate that our method significantly outperforms existing RAG methods in both in-domain and out-of-domain settings, validating its effectiveness in complex multi-hop reasoning tasks.

In recent years, there has been a surge of interests in interpretable graph reasoning methods. However, these models often suffer from limited performance when working on sparse and incomplete graphs, due to the lack of evidential paths that can reach target entities. Here we study open knowledge graph reasoning---a task that aims to reason for missing facts over a graph augmented by a background text corpus. A key challenge of the task is to filter out "irrelevant" facts extracted from corpus, in order to maintain an effective search space during path inference. We propose a novel reinforcement learning framework to train two collaborative agents jointly, i.e., a multi-hop graph reasoner and a fact extractor. The fact extraction agent generates fact triples from corpora to enrich the graph on the fly; while the reasoning agent provides feedback to the fact extractor and guides it towards promoting facts that are helpful for the interpretable reasoning. Experiments on two public datasets demonstrate the effectiveness of the proposed approach. Source code and datasets used in this paper can be downloaded at https://github.com/shanzhenren/CPL

The general public shows great potential for utilizing scientific research. For example, a singer discovered her ectopic pregnancy by looking up clinical case reports. However, an exorbitant paywall impedes the public’s access to scientific literature. Our case study on a social network demonstrates a growing need for non-open access publications, especially for biomedical literature. The challenge is that non-open access papers are protected by copyright licenses that bar free distribution. In this paper, we propose a technical framework that leverages the doctrine of "idea-expression dichotomy" to bring ideas across paywalls. Idea-expression dichotomy prevents copyright holders from monopolizing ideas, theories, facts, and concepts. Therefore facts may pass through paywalls unencumbered by copyright license restrictions. Existing fact extraction methods (such as information extraction) require either large training sets or domain knowledge, which is intractable for the diverse biomedical scope spanning from clinical findings to genomics. We therefore develop a rule-based system to represent and extract facts. Social networkers and academics validated the effectiveness of our approach. 7 out of 9 users rated the paper’s information from the facts to be above average (≥6/10). Only 7% of the extracted facts were rated misleading.