I am Junhao Shen (Chinese name: 沈君豪), a senior undergraduate student in the School of Computer Science and Technology at East China Normal University. Currently, I conduct the research on the general planning capability of LLM under the supervision of Dr. Kai Chen in OpenMMLab at Shanghai Artificial Intelligence Laboratory.
I am interested in large language models. Previously, I also conduct the research regarding interpretable machine learning for intelligent education and graph representation learning.
I am expected to graduate from East Chine Normal University in June 2025, join the School of Artificial Intelligence at Shanghai Jiao Tong University as a Ph.D. student. If you would like to discuss further, please feel free to contact me via email, and I will respond as soon as possible.
Recent Updates
📰 Itinerary Notice
Speech
Opening Ceremony (Undergraduate Student Representative)
Time: UTC+8 13:30, September 13, 2024
Venue: Stadium 2F, East China Normal University (Putuo Campus), Shanghai, China
💻 Academic Progress
[July 31, 2024] The Modularized Python Toolkit "InsCD" for Cognitive Diagnosis Model is Now Released [Project Page].
[June 13, 2024] Awarded SenseTime Scholarship Nomination (Top 30).
[May 17, 2024] One Paper Accepted by KDD 2024 and Scheduled for Oral Presentation in Barcelona [Project Page].
More Information
[May 17, 2024] Champion of the Undergraduate Academic Show of China Computer Federation.
[April 1, 2024] Awarded Nezha Technology Outstanding Student Excellence Scholarship.
[January 23, 2024] One Paper Accepted by WWW 2024.
[December 20, 2023] Gold Prize for the NeurIPS 2023 Causal Structure Learning Competition.
[December 9, 2023] One Paper Accepted by AAAI'24 and Scheduled for Presentation in Vancouver. [Project Page]
[October 1, 2023] Awarded Undergraduate First-Class Scholarship.
[April 7, 2023] Granted a Full Exchange Scholarship by City University of Hong Kong and Scheduled to Go on Exchange in August (2023).
[September 30, 2022] Awarded Undergraduate Special Scholarship.
[September 7, 2022] Qualified to Teach at "Top-class Course" and Scheduled to Teach C Programming Language Next Month (October).
Publications
*:Corresponding Author #:Equal Contribution
Capturing Homogeneous Influence among Students: Hypergraph Cognitive Diagnosis for Intelligent Education Systems
Junhao Shen, Hong Qian*, Shuo Liu, Wei Zhang, Bo Jiang, Aimin Zhou. In Proceedings of the 30th SIGKDD Conference on Knowledge Discovery and Data Mining (KDD'24), Barcelona, Spain.
Oral Presentation
Symbolic Cognitive Diagnosis via Hybrid Optimization for Intelligent Education Systems
Junhao Shen, Hong Qian*, Wei Zhang, Aimin Zhou. In Proceedings of the 38th AAAI Conference on Artificial Intelligence (AAAI'24), Vancouver, Canada, 14928-14936.
As the only student author
More Papers
Shuo Liu#, Junhao Shen#, Hong Qian*, Aimin Zhou. Inductive Cognitive Diagnosis for Fast Student Learning in Web-based Online Intelligent Education Systems. In Proceedings of the ACM Web Conference 2024 (WWW'24), Singapore, Singapore. [PDF] [Code]
[Under Review] Hong Qian*, Mingjia Li, Shuo Liu, Junhao Shen, Wei Zhang, Qi Liu and Aimin Zhou. A Causal Decision-Making Framework for Long-Term Cumulative Cognitive Enhancement of Personalized Student Learning. IEEE Transactions on Learning Technologies.
Research Experience
OpenMMLab, Shanghai Artificial Intelligence Laboratory, Shanghai, China
Research Internship from August, 2024 to Present
Research Field: general planning capability of LLM
Advisor: Dr. Kai Chen
The University of Hong Kong, Hong Kong S.A.R., China
Visiting Student from July 2023 to August 2023
ECNU-HKU Cooperation Project, Granted a Full Visiting Scholarship
Visiting Content: graph representation learning and graph neural networks
Shanghai Institute of AI Education, Shanghai, China
Research Internship from July, 2022 to July, 2024
Research Field: interpretable machine learning for intelligent education.
Advisors: Assoc. Prof. Hong Qian, Prof. Wei Zhang and Prof. Aimin Zhou
Achievements: Published three papers as the first author/co-author in CCF-A international conferences.
Education
East China Normal University, Shanghai, China
From September 2021 to Present
Undergraduate, Computer Science and Technology
Advisors: Assoc. Prof. Hong Qian, Prof. Wei Zhang and Prof. Aimin Zhou.
City University of Hong Kong, Hong Kong S.A.R., China
From August 2023 to December 2023
Exchange Student, Computer Science, Granted a Full Exchange Scholarship
High School Affiliated to Southwest University, Chongqing, China
From September 2018 to June 2021
Senior High Student, Class 1, Cohort 2021
Award and Achievement
[June, 2024] SenseTime Scholarship Nomination
Top 30 among all undergraduate students majored in computer science in Mainland China.
[May, 2024] Champion of the Undergraduate Academic Show of China Computer Federation
Top 1 among all undergraduate participants majored in computer science in Mainland China.
[April, 2024] Nezha Technology Outstanding Student Excellence Scholarship
Top 2 among the school of computer science and technology at East China Normal University.
[December, 2023] Gold Prize for NeurIPS 2023 Causal Structure Learning Competition
Top 1 among hundreds of teams from all over the world.
[October, 2023] Undergraduate First-Class Scholarship
Top 10% among undergraduate student in the school of computer science and technology at East China Normal University.
[September, 2022] Undergraduate Special Scholarship
Top 3% among undergraduate student in the school of computer science and technology at East China Normal University.
Speech & Talk
Capturing Homogeneous Influence among Students: Hypergraph Cognitive Diagnosis for Intelligent Education Systems
Time: UTC+8 13:30 - 13:50, August 14, 2024
Venue: AI TIME, Virtual
Time: UTC+2 12:00 - 12:15, August 29, 2024
Venue: Room 134, Centre de Convencions Internacional de Barcelona, Spain
Cognitive diagnosis aims to infer the students' proficiency levels on each knowledge concept. We observes that most existing methods can hardly effectively capture the homogeneous influence due to its inherent complexity, resulting in shortcomings in interpretability and efficacy. In this talk, we introduce a hypergraph cognitive diagnosis model (HyperCDM) to effectively capture the homogeneous influence. Extensive experiments on both offline and online real-world datasets show that HyperCDM achieves state-of-the-art performance in terms of interpretability and capturing homogeneous influence effectively.
Interpretable Symbolic Cognitive Diagnosis Method in Intelligent Education Systems
Time: UTC+8 16:00 - 16:20, May 17, 2024
Venue: Young Elite Forum of China Computer Federation, Ningbo, China
Cognitive diagnosis is a fundamental upstream task in intelligent education. In the real-world scenarios, students and teachers require that models' training process, inference and output are interpretable. In this talk, we introduce the new method that incorporate symbolic regression into cognitive diagnosis to enhance the interpretability.
Academic Service
[From October, 2022 to November 2022] Lecturer of "Top-class Course" C Programming Language.