I am currently at the Artificial Intelligence Research Institute of Soochow University, advised by Assoc. Prof. Juntao Li(李俊涛) and Prof. Min Zhang(张民).
Before this, I received my Bachelor’s degree (2018-2022, computer science) from Soochow University.
I am currently working as a research intern at Shanghai AI Laboratory General Technology Research under the mentorship of Xiaoye Qu(瞿晓晔) and Yu Cheng(成宇).
🤔 My research interests are primarily focused on exploring various challenges related to the dynamic of “time” in machine learning. Currently, my main focus is on Large Language Models (LLMs), covering the following topics:
- Adaptation and Generalization: Focusing on the reasons behind model degradation in time and exploring the strategies to improve temporal generalization and update the knowledge without catastrophic forgetting of old knowledge.
- Trustworthiness and Reliability: Investigation into the behavior of LLMs when encountering knowledge conflicts and exploring methods to mitigate knowledge conflicts of old knowledge as we update LLMs with new knowledge.
- Complex reasoning: Research on equipping LLMs with human-level temporal reasoning and commonsense. This involves basic temporal concepts understanding and intricate temporal interpretations and computations.
📖 Can the progress of civilization, and the advancement of technology truly transcend time? Or are we just endlessly pursuing the elusive dust in the hourglass?
🤝 I’m looking for a PhD position in 2025 Fall. Don’t hesitate to email me if there is a potential opportunity!
🔥 News
- 2024.7: 🎉🎉 One paper is accepted by COLM 2024!
- 2024.6: 🎉🎉 Invited talk on temporal reasoning at NICE!
- 2024.5: 🎉🎉 One paper is accepted by ACL 2024 oral !
- 2023.10: 🎉🎉 One paper is accepted by EMNLP 2023!
- 2022.10: 🎉🎉 One paper is accepted by EMNLP 2022!
📝 Publications
-
COLM 2024
Timo: Towards Better Temporal Reasoning for Language Models, Zhaochen Su, Jun Zhang, Tong Zhu, Xiaoye Qu, Juntao Li, Min Zhang, Yu Cheng. -
ACL 2024 oral
Living in the Moment: Can Large Language Models Grasp Co-Temporal Reasoning?, Zhaochen Su, Juntao Li, Jun Zhang, Tong Zhu, Xiaoye Qu, Pan Zhou, Yan Bowen, Yu Cheng, Min zhang -
EMNLP 2023
Efficient Continue Training of Temporal Language Model with Structural Information, Zhaochen Su, Juntao Li, Zikang Zhang, Zihan Zhou, Min Zhang -
EMNLP 2022
Improving Temporal Generalization of Pre-trained Language Models with Lexical Semantic Change, Zhaochen Su, Zecheng Tang, Xinyan Guan, Lijun Wu, Min Zhang, Juntao Li -
Submission to Neurips-24
ConflictBank: A Benchmark for Evaluating Knowledge Conflicts in Large Language Models, Zhaochen Su, Jun Zhang, Xiaoye Qu, Tong Zhu, Yanshu Li, Jiashuo Sun, Juntao Li, Min Zhang, Yu Cheng. -
Submission to AI-Journal
Language Model Adaptation as Lexical Semantic Change Modeling, Zhaochen Su, Zikang Zhang, Juntao Li, Lijun Wu, Zecheng Tang, Min Zhang.
🎖 Honors and Awards
- Soochow University Graduate Outstanding Scholarship (Rank 1st)
- The Samsung Scholarship
- China Software Design Competition – First Prize (National Level, Team Leader)
- Chinese Collegiate Computing Competition – Second Prize (National Level, Team Leader)
📖 Educations
- 2022.09 - current, Second-year Master, Artificial Intelligence Research Institute, Soochow University, Suzhou.
- 2018.09 - 2022.06, Bachelor, Institute of Computer Science and Technology, Soochow University, Suzhou.
- 2021.07 - 2022.07, Dual Degree, Brain Science, Soochow University, Suzhou.
💁 Volunteer
- Conference Reviewer: ARR, ACL-24, EMNLP-24, MM-24, ICLR-25
- 2023.09 - current, Soochow University Graduate Student Union - President
- 2020.05 - 2021.05, Students’ International Communication Association (SICA) - President
💻 Internships
- 2023.10 - current, Shanghai AI Laboratory, China.
- 2021.06 - 2021.09, World Heritage Institute of Training and Research for the Asia and the Pacific Region under the auspices of UNESCO, China.
🎨 Miscellaneous
Without any special circumstances, I exercise for 1 hour every day. Here are some of my favorite sports and personal achievements: 🏋️ fitness (Progressive Overload (PR): bench press 98.5 kg, squat 140 kg, deadlift 135 kg), 🏓 table tennis (amateur intermediate), 🎾 tennis (skill level: 2.5), 🏃 running (5 km/h, 10 km pace), 🚴 cycling (25.5 km/h, 50 km pace).
If you share similar hobbies, please feel free to reach out and connect!