Mohammadamin (Amin) Banayeeanzade


Profile picture

I'm a fourth-year Ph.D. student in Computer Science at the University of Southern California (USC), advised by Sai Praneeth Karimireddy. I study the trustworthiness of AI systems with a particular focus on diversity in large language models (LLMs) and agentic systems. My research investigates the diversity problem in current models, which often produce homogeneous outputs, follow narrow exploration traces, and rely on limited or biased inputs during decision-making. By systematically characterizing and improving diversity in these aspects, my work aims to enhance the robustness, fairness, and reliability of AI systems, ultimately strengthening their trustworthiness in real-world deployment.

Prior to USC, I received my Master's at Computer Engineering department of Sharif University of Technology under the supervision of Mahdieh Soleymani, where I did research on developing brain-inspired algorithms for the meta-continual learning problem. Additionally, I got my B.Sc. in Electrical Engineering from the same university where I was jointly supervised by Mahdi Shabany and Zahra Kavehvash.

Mail Twitter Scholar Github LinkedIn

Publications

Sparks of Rationality: Do Reasoning LLMs Align with Human Judgment and Choice?
A. N. Tak *, A. Banayeeanzade*, A. Bolourani *, F. Bahrani, A. Chaubey, S. P. Karimireddy, N. Schwarz, J. Gratch
underreview at ICML 2026.
[Paper] [Code]

EPSVec: Efficient and Private Synthetic Data Generation via Dataset Vectors
A. Banayeeanzade*, Q. Yang*, D. Fu, S. Hong, E. Babinsky, A. Samuel, A. Kumar, R. Jia, S. P. Karimireddy
underreview at ICML 2026.
[Paper] [Code]

AutoFocus-IL: VLM-based Saliency Maps for Data-Efficient Visual Imitation Learning without Extra Human Annotations
L. Gong, F. Bahrani, Y. Zhou, A Banayeeanzade, J. Li, E. Bıyık,
ICRA, 2026.
[Paper] [Project Page]

Psychological Steering in LLMs: An Evaluation of Effectiveness and Trustworthiness
A. Banayeeanzade*, A. N. Tak*, F. Bahrani, A. Bolourani, L. Blas, E. Ferrara, J. Gratch, S. P. Karimireddy
arXiv, 2025.
[Paper] [Code]

Mechanistic Interpretability of Emotion Inference in Large Language Models
A. N. Tak*, A. Banayeeanzade*, A. Bolourani, M. Kian, R. Jia, J. Gratch.
ACL, 2025.
[Paper] [Code]

Theoretical Insights into Overparameterized Models in Multi-Task and Replay-Based Continual Learning
A. Banayeeanzade, M. Soltanolkotabi, M. Rostami.
TMLR, 2025.
[Paper] [Code]

GABRIL: Gaze-Based Regularization for Mitigating Causal Confusion in Imitation Learning
A. Banayeeanzade*, F. Bahrani*, Y. Zhou, E. Bıyık.
IROS, 2025.
[Paper] [Project Page]

Hybrid Learners Do Not Forget: A Brain-Inspired Neuro-Symbolic Approach to Continual Learning
A. Banayeeanzade, M. Rostami.
arXiv, 2025.
[Paper]

A Distinct Unsupervised Reference Model from The Environment Helps Continual Learning
A. Ameli, A. Banayeeanzade, M. Samiei, M. Soleymani.
arXiv, 2023.
[Paper]

Generative vs. Discriminative: Rethinking The Meta-Continual Learning
A. Banayeeanzade*, R. Mirzaiezadeh*, H. Hasani, M. Soleymani.
NeurIPS, 2021.
[Paper] [Code]
* Equal Contributions.

Other Projects

Automatic Object Detection Under Clothing in Millimeter-Wave Images
Accomplished project as senior AI researcher at Basir Wave Tech
[Video]