Search

Large Language Models (LLMs) have demonstrated impressive capabilities across a wide range of tasks. However, their proficiency and reliability in the specialized domain of Data Analysis, particularly with a focus on data-driven thinking, remain uncertain. To bridge this gap, we introduce BIBench, a comprehensive benchmark designed to evaluate the data analysis capabilities of LLMs within the context of Business Intelligence (BI). BIBench assesses LLMs across three dimensions: 1) BI foundational knowledge, evaluating the models' numerical reasoning and familiarity with financial concepts; 2) BI knowledge application, determining the models' ability to quickly comprehend textual information and generate analysis questions from multiple views; and 3) BI technical skills, examining the models' use of technical knowledge to address real-world data analysis challenges. BIBench comprises 11 sub-tasks, spanning three categories of task types: classification, extraction, and generation. Additionally, we've developed BIChat, a domain-specific dataset with over a million data points, to fine-tune LLMs. We will release BIBenchmark, BIChat, and the evaluation scripts at \url{https://github.com/cubenlp/BIBench}. This benchmark aims to provide a measure for in-depth analysis of LLM abilities and foster the advancement of LLMs in the field of data analysis.

We investigate the ramification group filtration of a Galois extension of function fields, if the Galois group satisfies a certain intersection property. For finite groups, this property is implied by having only elementary abelian Sylow p-subgroups. Note that such groups could be non-abelian. We show how the problem can be reduced to the totally wild ramified case on a p-extension. Our methodology is based on an intimate relationship between the ramification groups of the field extension and those of all degree p sub-extensions. Not only do we confirm that the Hasse-Arf property holds in this setting, but we also prove that the Hasse-Arf divisibility result is the best possible by explicit calculations of the quotients, which are expressed in terms of the different exponents of all those degree p sub-extensions.

Wei-Wei Tang,1,* Qingquan Wu,3,* Su-Qing Li,1 Yu-Suo Tong,1 Zi-Hao Liu,1 Tong-Xin Yang,1 Yong Xu,1,2 Xiu-Feng Cao1 1Department of Surgical Oncology, Nanjing First Hospital, 2Cancer Center, Nanjing Medical University, Nanjing, People’s Republic of China; 3Department of Cardiothoracic Surgery, Huaian No 1 People’s Hospital, Affiliated to Nanjing Medical University, Nanjing, People’s Republic of China *Indicates both are first authors Abstract: Long non-coding RNAs (lncRNAs), transcripts as longer than 200 nt in length with a great number of varieties in human genomics, play important roles in the regulation of genetics and epigenetics including gene transcription and post-transcription. Increasing evidence have demonstrated the upregulation of lncRNAs in tumorigenesis and metastasis of esophageal cancer (EC), a type of malignant tumors particularly in Asia. In this review, we briefly discuss the profiles and functions of lncRNAs involved in the progression of EC, which may provide a new approach to improve EC diagnosis and treatment. Keywords: long non-coding RNAs, esophageal cancer, biomarker, review