Alireza Ziabari

Alireza Ziabari

NLP at USC

Helloo

My name is Alireza Ziabari. I am a fifth year CS Ph.D. student at University of Southern California, advised by Morteza Dehghani. Previously, I got my B.S. in Computer Engineering from Sharif University of Technology, where I worked with Mohammad Hossein Rohban.

I am broadly interested in developing adaptive NLP systems that can learn and reason in ways comparable to humans. My work explores reinforcement learning (RL) as a mechanism for improving Large Language Models (LLMs), focusing on how RL can guide reasoning processes.

Publications

  1. A primer for evaluating large language models in social-science research Suhaib Abdurahman, Alireza S. Ziabari, Alexander Moore, Daniel Bartels, Morteza Dehghani. AMPPS.
  2. Cost-Efficient Subjective Task Annotation and Modeling through Few-Shot Annotator Adaptation Preni Golazizian, Alireza S. Ziabari, Ali Omrani, Morteza Dehghani. EMNLP 2024.
  3. Evaluating Creativity and Deception in Large Language Models: A Simulation Framework for Multi-Agent Balderdash Parsa Hejabi, Elnaz Rahmati, Alireza S. Ziabari, Preni Golazizian, Jesse Thomason, Morteza Dehghani. Wordplay at ACL 2024.
  4. Reinforced Multiple Instance Selection for Speaker Attribute Prediction Alireza S. Ziabari, Ali Omrani, Parsa Hejabi, Preni Golazizian, Payam Piray, Morteza Dehghani. NAACL 2024.
  5. Towards a Unified Framework for Adaptable Problematic Content Detection via Continual Learning Ali Omrani*, Alireza S. Ziabari*, Preni Golazizian, Jeffery Sorensen, Morteza Dehghani. WOAH at NAACL 2024.
  6. Social-group-agnostic bias mitigation via the stereotype content model Ali Omrani, Alireza S. Ziabari, Charles Yu, Preni Golazizian, Brendan Kennedy, Mohammad Atari, Heng Ji, Morteza Dehghani. ACL2023.

Preprint

  1. The Homogenizing Effect of Large Language Models on Human Expression and Thought.
  2. Reasoning on a Spectrum: Aligning LLMs to System 1 and System 2 Thinking.
  3. CoCo-CoLa: Evaluating Language Adherence in Multilingual LLMs.
  4. The Shrinking Landscape of Linguistic Diversity in the Age of Large Language Models.
  5. The Moral Foundations Reddit Corpus.
  6. What Doesn’t Kill Us: Harsh Ecologies Beget Difficulty-as-Improvement Norms.

Work Experiences

Netflix, Research Intern. May 2025 - Aug. 2025

Tapsi, Machine Learning Enigneer. Oct 2020 - May 2021

Fanap, Machine Learning Intern. Jul 2020 - Oct 2020

Miscellaneous

Soccer and Rock Climbing!