Research: LLM Reasoning and Auto-Evaluation, Fintech, Embodied AI
I am very interested in the reasoning and auto-evaluation of large language models, aiming to enable iterative evolution of LLMs and enable comprehensive evaluation in a more automated way. At the same time, I have also done some fintech work in the past.
- LLM Reasoning and Auto-Evaluation: Utilize LLMs to reason about complex problems such as code and mathematics, and use automated new methods to solve the problems of data leakage and difficulty in generating evaluations encountered during the model evaluation process.
- Fintech: Using technologies such as LLMs, graph models, and time-series models to solve problems in the financial field such as intelligent investment advisory, intelligent text generation, and risk control.
- Embodied AI: Using multi-modal models to manipulate, navigate, and enhance the emotional interaction capabilities of robots or mechanical dogs.
|
|
R3-NL2GQL: A Model Coordination and Knowledge Graph Alignment Approach for NL2GQL
Yuhang Zhou,
Yu He,
Siyu Tian,
Yuchen Ni,
Zhangyue Yin,
Xiang Liu,
Chuanjun Ji,
Sen Liu,
Xipeng Qiu,
Guangnan Ye,
Hongfeng Chai
Findings of EMNLP, 2024   (Spotlight of IJCAI 2024 Workshop)
arXiv
|
|
SilverSight: A Multi-Task Chinese Financial Large Language Model Based on Adaptive Semantic Space Learning
Yuhang Zhou,
Zeping Li,
Siyu Tian,
Yuchen Ni,
Sen Liu,
Guangnan Ye,
Hongfeng Chai
CCL / Journal of Chinese Information Processing, 2024  
arXiv
|
|
Are Large Language Models Rational Investors?
Yuhang Zhou,
Yuchen Ni,
Zhiheng Xi,
Zhangyue Yin,
Xiang Liu,
Jian Zhang,
Sen Liu,
Yixin Cao,
Xipeng Qiu,
Guangnan Ye,
Hongfeng Chai
Under-Review, 2024  
arXiv
|
|
Research on Improving the Performance of Tri-ellipsoid Multi component Optical Sensors
Yuhang Zhou,
Yuchen Ni,
Wei Ge,
Ya Guo
Laser & Optoelectronics Progress, 2023  
arXiv
|
|
ChatReport: A Financial Research Report Generation Approach Based on Internet of Agents
Yuhang Zhou,
Hongjie Xia,
Yunhui Gan,
Yuchen Ni,
Siyu Tian,
Jiahui Zhang,
Yu Liu,
Guangnan Ye
Using multi-agent collaboration technology based on LLMs to complete the writing of financial multimodal long texts (such as financial research reports, bond applications, etc.), using a combination of large and small models to extract causal relationships from past research report data, simulating multi role interactions to form a document writing team, and achieving low-cost, long text financial document writing. The project has been implemented and applied in companies such as Guotai Junan Securities, DataGrand, and Baidu.
National 1st Prize, the Graduate Financial Technology Innovation Competition, 2023
Top 10 Excellence Awards, the 2024 LLM Financial Application Innovation and Practice Competition, 2024
|
-
Great Wall Fund - Financial quantification - Research Intern (Jul. 2023 - Nov. 2023)
-
Wind Inc. - Financial quantification - Research Intern (Feb. 2023 - Jul. 2023)
-
Huachuang Securities - Financial NLP tasks - Research Intern (Dec 2022 - Jan. 2023)
|
-
National 1st Prize, the Graduate Financial Technology Innovation Competition, 2023
-
First Class Academic Scholarship, Fudan University , 2022~2024
-
Outstanding Graduation Program(recommended for Provincial Outstanding), Jiangnan University, 2022
-
National 2nd Prize, The Blue Bridge Cup Software and Information Technology Competition, 2022
-
Huawei Excellent Scholarship, Jiangsu, 2022
-
Outstanding Youth League Cadres, Jiangnan University, 2022
-
Academic Scholarship, Jiangnan University, 2019~2022
|
|