Tianyi Niu

Hello! I am pursuing a Master of Science in Computer Science at The University of North Carolina at Chapel Hill. I am currently a Research Assistant at MURGe-Lab, advised by Prof. Mohit Bansal and Prof. Elias Stengel-Eskin. I am applying to PhD programs starting Fall 2026.

My current research focuses on two key directions:

  1. Abstractions and Cognition of MLLMs. How do abstractions formed by (M)LLMs differ from humans, and can insights from human cognition inform model development, and vice versa?
  2. Interaction and communication in multi-agent systems. How is knowledge transferred in multi-agent systems, and can principles of human communication also inform and optimize multi-agent interaction paradigms?

I previously graduated from UNC Chapel Hill highest distinction, obtaining a BS in Computer Science with highest honors and a BA in Linguistics. My undergraduate thesis (advised by Prof. Shashank Srivastava) explored whether LLMs benefit from curriculum training on synthetic pre-training data.

Feel free to reach me at: tianyin4 [at] cs [dot] unc [dot] edu

📢 News

Aug 21, 2025
New preprint: RotBench: Evaluating Multimodal Large Language Models on Identifying Image Rotation.
Aug 20, 2025
1 paper accepted into EMNLP 2025: Chameleon LLMs: User Personas Influence ChatBot Personality Shifts.
June 01, 2025
Joined MURGe-Lab as a Research Assistant advised by Prof. Mohit Bansal.

📝 Publications

Preprint
Socratic Students: Teaching Language Models to Learn by Asking Questions
Rajeev Bhatt Ambati, Tianyi Niu*, Aashu Singh, Shlok Mishra, Shashank Srivastava, Snigdha Chaturvedi
Arxiv preprint
2025
Preprint
Adversarially Probing Cross-Family Sound Symbolism in 27 Languages
Anika Sharma, Tianyi Niu, Emma Wrenn, Shashank Srivastava
Arxiv preprint
2025
Preprint
RotBench: Evaluating Multimodal Large Language Models on Identifying Image Rotation
Tianyi Niu, Jaemin Cho, Elias Stengel-Eskin, Mohit Bansal
Arxiv preprint
2025
EMNLP
Chameleon LMMs: User Personas Influence Chatbot Personality Shifts
Jane Xing, Tianyi Niu, Shashank Srivastava
The 2025 Conference on Empirical Methods in Natural Language Processing
2025
ICIMTH
Performance Evaluation of Popular Open-Source Large Language Models in Healthcare
Saif Khairat, Tianyi Niu, John Geracitano, Kaushalya Mendia
International Conference in Greece on Informatics, Management and Technology in Healthcare 2025
2025
ICIMTH
Acceptability of Academic Large Language Model for Patients Seeking Health Information
Saif Khairat, John Geracitano, Tianyi Niu, Kaushalya Mendia
International Conference in Greece on Informatics, Management and Technology in Healthcare 2025
2025
sigmorphon
Probing Neural Network Generalization using Default Patterns
Brandom Prickett*, Tianyi Niu*, Katya Pertsova
The 22nd SIGMORPHON workshop on Computational Morphology, Phonology, and Phonetics
2025