Wenqing Hou, Hongkui Tu, Ye Wang, Yue Zhang, Yuying Liu, Dong Zhu, Liqun Gao, Bin Zhou. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2026.

When fact-checking methods based on large language models (LLMs) use external evidence to validate claims, knowledge conflicts often arise. These conflicts typically stem from inconsistencies between the external evidence and LLMs' internal pre-existing knowledge. Such an inconsistency could lead LLMs to draw incorrect answers when validating claims, especially when they are overly confident in their internal incorrect knowledge. Previous works on LLM-based fact-checking have overlooked this issue. This paper, for the first time, proposes a framework (namely KnowFC) to navigate this issue. Our key insight is dividing and adaptively utilizing the knowledge that LLMs know and do not know, thereby avoiding conflicts while enhancing the correctness and efficiency of fact-checking. Specifically, in KnowFC, we propose an adaptive retrieval method, where we train an LLM using a reinforcement learning algorithm coupled with the Dunning-Kruger effect-inspired reward mechanism to identify its knowledge boundaries through confidence calibration, thereby realizing adaptive evidence retrieval. Besides, we propose a reliable and debiased fact verification method, where we organize and construct reasoning graphs using retrieved evidence to verify claims, followed by a causal intervention method using causal mediation analysis to mitigate internal knowledge interference. Experimental results on both FEVEROUS and AVeriTeC datasets show that our method outperforms baseline methods in terms of accuracy and F1 score, while also improving fact-checking efficiency.

The goal of terrain viewshed point selection is to identify multiple viewpoints within a specific area that offer optimal visibility. However, as the resolution of terrain data becomes finer, the number of data points to be processed grows significantly, leading to a sharp rise in computational demands for viewshed point selection. This paper introduces the Efficient Precise Viewshed Point Selection (EPVPS) algorithm, which provides an important reference for multi-viewpoint planning. First, the Empty Circles based K-means (ECKM) algorithm is utilized to determine the initial centers for candidate viewpoint clustering. Second, a new viewpoint evaluation metric, Weighted Coverage Overlap Metric (WCOM), is proposed. This metric divides the viewshed of candidate points into coverage contribution points and overlap contribution points, computes WCOM based on this division, and stores the values in a min-heap. Finally, viewpoints are added and deleted from clusters by adjusting the sets of coverage contribution points and overlap contribution points. The EPVPS algorithm is compared with the Candidate Viewpoints Filtering (CVF) algorithm and the Simulated Annealing (SA) algorithm. Experimental results demonstrate that the EPVPS algorithm outperforms the others in computational efficiency, coverage rate, and overlap rate.