Skip to content

Ms-March7/academic-profile

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

1 Commit
 
 

Repository files navigation

Beichen Wang (王北辰)

Ph.D. student in Computational and Applied Mathematics at Qiuzhen College, Tsinghua University (2026.09-2029.06).

I work at the intersection of Transformer-based large language models, quantitative research, computational mathematics, and AI for Science. My current interests include autoregressive modeling, LLM agents, event-driven market signals, and research workflow automation.

Contact: wbc23@mails.tsinghua.edu.cn


Research Interests

  • Transformer-based and autoregressive models: large language models, tool use, retrieval, agents, and structured reasoning workflows.
  • LLM-assisted quantitative research: event-driven signal extraction, news-to-signal pipelines, time-series modeling, and risk-aware validation.
  • Computational mathematics: numerical methods, scientific computing, PDEs, multiscale modeling, and homogenization theory.
  • AI for Science: AI-assisted mathematical discovery, molecular/material design, and research automation.

Publications & Preprints


Selected Work

AI-Assisted Mathematical Discovery

Co-first author of AI Mathematician as a Partner in Advancing Mathematical Discovery, an ICAIS 2025 Outstanding Paper.

  • Focus: LLM-assisted proof exploration, human-AI collaboration, and mathematical discovery workflows.
  • Link: arXiv:2510.26380

LLM-Based Quantitative Research

Working on a private AI-for-Trading research prototype with live-market validation and investor-facing exploration.

  • Focus: event-driven market signals, news-to-signal extraction, structured research workflows, and realistic validation.
  • Status: private prototype; implementation details and performance metrics are not publicly disclosed.

Computational Mathematics Background

Earlier mathematical research focused on PDEs and quantitative homogenization theory, especially the Lamé-Stokes coupled system.

  • Link: arXiv:2606.05098
  • Topics: two-scale expansions, mixed variational methods, convergence-rate estimates, and corrector regularity.
  • This work trained me in rigorous analysis, multiscale modeling, and proof-based problem solving.

AI for Molecules and Materials

I have also explored AI for Science problems related to molecular and material design, including OLED molecular design and MOF representation/generation.


Education

  • Ph.D. in Computational and Applied Mathematics, Qiuzhen College, Tsinghua University, 2026.09-2029.06
  • M.S. in Mathematics, Qiuzhen College, Tsinghua University, 2023.09-2026.06
  • B.S. in Mathematics and Applied Mathematics, Ningbo University, 2019.09-2023.06

Skills

  • Programming: Python, C/C++
  • LLM systems: Transformer-based models, autoregressive modeling, prompt workflows, tool use, retrieval, and structured extraction
  • Quantitative research: event-driven signals, market data workflows, time-series modeling, backtesting concepts, and risk-aware evaluation
  • Mathematics: numerical analysis, PDEs, multiscale modeling, probability, statistics, and scientific computing

Honors

  • ICAIS 2025 Outstanding Paper Award
  • Tsinghua Qiuzhen College Creative Paper Award, 2025
  • Tsinghua University Graduate with Merit, 2026
  • Chinese Mathematics Competitions, Final Round, Second Prize, 2023
  • Mathematical Contest in Modeling, Honorable Mention
  • Shenzhen Cup Mathematical Modeling Challenge, Third Prize
  • Zhejiang Provincial Government Scholarship, 2020, 2021, 2022

Open To

Research internships and collaborations in quantitative research, LLM systems, computational mathematics, and AI for Science.

About

Academic profile of Beichen Wang: computational mathematics, LLMs, quantitative research, and AI for Science.

Topics

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

No releases published

Packages

 
 
 

Contributors