CV
General Information
| Full Name | He, Pengfei |
Education
-
2022-present
Computer Science PhD
Department of Computer Science and Engineering, Michigan State University, Michigan, USA
-
2020-present
Statistics PhD
Department of Statistics and Probability, Michigan State University, Michigan, USA
-
2019-2020
Statistics M.S.
Department of Statistics, University of Wisconsin, Madison, Wisconsin, USA
-
2015-2019
Mathematics B.S.
Department of Statistics, Nankai University, Tianjin, China
Experiences
-
2025 Fall & 2026 Spring
- Student researcher at CloudAI Research, Google
-
2025 Spring & Summer
- Applied Scientist Intern at Amazon.Inc (Core search & AWS Ads)
-
2024 Summer
- Research Intern at Alibaba Group (US)
-
2023-present
- Visiting research student at Okinawa Institute of Science and Technology (OIST)
-
2022
- Research Scientist at James Madison College, Michigan State university
Honors and Awards
-
2026
- Outstanding Graduate Student Award, MSU
-
2025
- ICLR 2025 Notable Reviewer
- SDM 2025 Doctoral Forum Travel Award
-
2023
- NeurIPS scholar award
-
2022
- CIKM-2022 Student Registration Award
-
2022
- KDD-2022 Student Registration Award
-
2021
- Awarded the Professor James Stapleton Prize in Statistics, Michigan State University
-
2016
- Recipient of Second Prize, Nankai University
Talks
-
2026
- Invited talk at University of Illinois Chicago, CS517
- The evolution of AI security
-
2026
- On-site talk at Lamar University
- The evolution of AI security
-
2026
- Virtual talk at University of Arizona
- The evolution of AI security
-
2025
- Virtual talk at University of Kentucky
- The evolution of AI security
-
2025
- Invited talk at University of Georgia
- The evolution of AI security
-
2025
- Invited talk on JSM 2025
- Understanding In-context-learning from the lens of Bayesian Statistic
-
2023
- Invited talk at AI-Time
- Probabilistic Categorical Adversarial Attack and Jailbreak
-
2023
- Oral presentation at Okinawa Institute of Science and Technology OIST
- Probabilistic Categorical Adversarial Attack
-
2022
- Oral presentation at CIKM 2022
- {"PROPN"=>"Personalized Probabilistic Strategic Parameter Optimization in Recommendations"}
Academic Interests
-
LLM Agents
- Enhance LLM-based agents functionality (reasoning, tool usage)
- Vulnerablities of LLM Agents (Privacy, Communication attacks, Multi-agent vulnerablities)
-
Truthworthy LLMs
- Data poisoning attacks
- Understanding of in-context learning and chain-of-thought reasoning
- Jailbreak attacks and interpretibility
-
Adversarial learning.
- Data poisoning attack and defense
- Model robustness and generalization
- Data poisoning
-
Machine learning & Statistics.
- Deep model expressiveness and overfitting
- High-dimensional statistics
-
Diffusion models.
- Watermark and backdoor attacks.
Services
- PC member/reviewer for AAAI, KDD, SDM, ICDM, PAKDD, WWW, ACL, EMNLP, ICML, NeruIPS, ICLR, AISTAT, TKDE, TKDD, TMLR, JASA.
- SPC for TKDD.
- Volunteer for KDD2022
Other Interests
- Gym&Muscle; Tennis; Basketball; Badminton; Hiking
- Anime; Manga; Rock&Classic, fan of Linkin Park, Led Zeppelin, Guns N' Roses, Lynyrd Skynyrd, Queens; Play Violin
- Video games!!! Fan of Battlefield, Monster Huner, Death Stranding and many more