Scholar
Zhecheng Sheng
Google Scholar ID: 6GSRIycAAAAJ
University of Minnesota, Twin Cities
Natural Language Processing
Statistical Machine Learning
Trustworthy AI
Causal Inference
Follow
Homepage
↗
Google Scholar
↗
Citations & Impact
All-time
Citations
156
H-index
6
i10-index
4
Publications
20
Co-authors
5
list available
Contact
Email
sheng136@umn.edu
CV
Open ↗
GitHub
Open ↗
LinkedIn
Open ↗
Publications
5 items
Toward Unifying Group Fairness Evaluation from a Sparsity Perspective
2025
Cited
0
Mitigating Confounding in Speech-Based Dementia Detection through Weight Masking
2025
Cited
0
"Is There Anything Else?'': Examining Administrator Influence on Linguistic Features from the Cookie Theft Picture Description Cognitive Test
2025
Cited
0
On the Sequence Evaluation based on Stochastic Processes
arXiv.org · 2024
Cited
0
BBScore: A Brownian Bridge Based Metric for Assessing Text Coherence
AAAI Conference on Artificial Intelligence · 2023
Cited
4
Resume (English only)
Academic Achievements
Paper accepted to EMNLP 2025 Main Conference (Aug 2025)
Paper accepted to ACL 2025 Main Conference (May 2025); proposed weight masking to mitigate confounding bias in fine-tuning
Paper accepted to Journal of Biomedical Informatics (Jun 2025)
Passed Ph.D. preliminary exam and advanced to candidacy (Apr 2025)
Paper accepted to CMCL workshop @ NAACL 2025 (Mar 2025)
Paper accepted to ACL 2024 Findings (May 2024)
Paper on text coherence modeling accepted to AAAI 2024 (Dec 2023)
One paper and one abstract accepted by AMIA Annual Symposium 2023 (Jul 2023)
Paper accepted to DialDoc workshop @ ACL 2023 (May 2023)
Paper accepted to DistShift workshop @ NeurIPS 2023 (Oct 2023)
Background
5th-year Ph.D. student in Health Data Science at the University of Minnesota
Member of the Cognitive AI Lab at UMN
Research interests include Natural Language Processing (NLP) and Trustworthy Machine Learning
Current projects focus on ensuring fairness and robustness of ML/DL models in healthcare applications
Addresses issues related to sensitive attributes and confounding distribution shifts affecting model performance
Also interested in probing Large Language Models (LLMs) to evaluate faithfulness of generation and adapting them for domain-specific tasks
Co-authors
5 total
Serguei Pakhomov
University of Minnesota
Trevor Cohen
University of Washington
Xiruo Ding
Stanford University
Changye Li
University of Washington
Tianhao Zhang
University of Minnesota
×
Welcome back
Sign in to Agora
Welcome back! Please sign in to continue.
Email address
Password
Forgot password?
Continue
Do not have an account?
Sign up