The 326th Wenlan Financial Forum

Abstract：

In recent years, the rapid advancement of artificial intelligence (AI) has driven innovation in large language models (LLMs). As one of the most cutting-edge technologies, LLMs are being widely applied across industries. For instance, DeepSeek, which gained widespread attention in 2023, represents one of the most advanced LLMs. It not only serves as a conversational agent but also demonstrates exceptional reasoning capabilities for complex problems, showcasing immense potential and application prospects.

The financial industry, which heavily relies on data analysis and information processing, has a strong demand for advanced AI technologies. Leveraging their powerful text comprehension, information extraction, reasoning, and predictive capabilities, LLMs hold significant value in various financial scenarios. Examples include: Assisting analysts in processing vast amounts of data and identifying market trends in investment analysis; Real-time monitoring of potential risk factors in risk management; Providing intelligent investment advice and financial planning recommendations in customer service; Automating document review and detecting anomalous transactions for compliance and regulatory purposes.As technology continues to evolve, the depth and breadth of LLM applications in finance will expand further, driving the industry toward greater intelligence and efficiency.

Speaker Introduction：

Lin Tao is a Financial Engineering Researcher at Guangfa Securities Co., Ltd. Development Research Center, holding a Master of Engineering degree. His research focuses on artificial intelligence, machine learning, and quantitative investment strategies.He has published first-authored papers at the China Conference on Pattern Recognition and Computer Vision (PRCV) and holds a computer software copyright (as the primary contributor). Additionally, he has won awards in multiple AI-related international and national competitions organized by institutions such as the National Natural Science Foundation of China and the International Conference on Computer Vision (ICCV).