Arshia Rafieioskouei

428 S Shaw Ln # 3320

East Lansing, MI 48824

Hi! I’m Arshia Rafieioskouei, a PhD student in Computer Science and Engineering at Michigan State University, working under the supervision of Dr. Borzoo Bonakdarpour in the TART Lab. I hold a Bachelor’s degree in Computer Engineering from Sharif University of Technology.

My research focuses on specification-based reinforcement learning, where I use logical specifications (such as temporal-logic formulas) to guide agents so that their learned policies satisfy high-level objectives and strict safety constraints. By encoding desired behaviors and forbidden scenarios as logical formulas, I can shape learning, constrain exploration, and verify that the resulting policies are safe and reliable in complex, real-world environments. In parallel, I work on causality, formalizing causal reasoning in temporal logic and translating it into a Satisfiability Modulo Theories (SMT) problem. This enables scalable causality analysis in Cyber-Physical Systems (CPS) to uncover causal relationships between events and understand why failures occur. I am also working on optimizing these techniques for better efficiency and exploring applications in medicine, nanoscience, and beyond.

Additionally, I’m actively seeking internships in fields such as RL, Safe Planning, Formal Methods, and CPS. You can learn more about my work and experience on my website or through my CV.

news

Apr 30, 2026	🤖 Our paper on HyPOLE: Hyperproperty-Guided Multi-Agent Reinforcement Learning under Partial Observation has been accepted to ICML 2026.
Oct 06, 2025	🚀 Our paper on Efficient Discovery of Actual Causality in Stochastic Systems has been accepted to VMCAI 2026.
Sep 18, 2025	🤖 Our paper on HYPRL: Reinforcement Learning of Control Policies for Hyperproperties has been accepted to NeurIPS 2025.
Apr 08, 2025	🧬 Our paper on Beyond Correlation: Establishing Causality in Protein Corona Formation for Nanomedicine has been accepted to Molecular Pharmaceutics.
Mar 01, 2025	🧬 Our paper on Ai-driven prediction of cardio-oncology biomarkers through protein corona analysis has been accepted to Chemical Engineering Journal.

selected publications

VMCAI’26

Efficient Discovery of Actual Causality in Stochastic Systems

Arshia Rafieioskouei^*, Kenneth Rogale^*, and Borzoo Bonakdarpour

Verification, Model Checking, and Abstract Interpretation, 2026

Bib PDF Code

@article{10.1007/978-3-032-15700-3_13,
  author = {Rafieioskouei, Arshia and Rogale, Kenneth and Bonakdarpour, Borzoo},
  title = {Efficient Discovery of Actual Causality in Stochastic Systems},
  journal = {Verification, Model Checking, and Abstract Interpretation},
  year = {2026},
  url = {https://link.springer.com/chapter/10.1007/978-3-032-15700-3_13#citeas},
  dimensions = {false},
  publisher = {Springer Nature Switzerland},
}

NeurIPS25

HYPRL: Reinforcement Learning of Control Policies for Hyperproperties

Tzu-Han Hsu^*, Arshia Rafieioskouei^*, and Borzoo Bonakdarpour

The Thirty-ninth Annual Conference on Neural Information Processing Systems, 2025

Bib PDF Poster Website

@article{Neurips25,
  title = {HYPRL: Reinforcement Learning of Control Policies for Hyperproperties},
  journal = {The Thirty-ninth Annual Conference on Neural Information Processing Systems},
  pages = {161134},
  year = {2025},
  dimensions = {false},
  url = {https://openreview.net/forum?id=lJSAtyx9Uc},
  author = {Hsu, Tzu-Han and Rafieioskouei, Arshia and Bonakdarpour, Borzoo},
}

Molecular Pharm

Beyond Correlation: Establishing Causality in Protein Corona Formation for Nanomedicine

Arshia Rafieioskouei, Kenneth Rogale, Amir Ata Saei, and 2 more authors

Molecular Pharmaceutics, Apr 2025

Bib PDF Code Website

@article{cite-key,
  annote = {doi: 10.1021/acs.molpharmaceut.5c00262},
  author = {Rafieioskouei, Arshia and Rogale, Kenneth and Saei, Amir Ata and Mahmoudi, Morteza and Bonakdarpour, Borzoo},
  date = {2025/04/09},
  doi = {10.1021/acs.molpharmaceut.5c00262},
  isbn = {1543-8384},
  journal = {Molecular Pharmaceutics},
  month = apr,
  publisher = {American Chemical Society},
  title = {Beyond Correlation: Establishing Causality in Protein Corona Formation for Nanomedicine},
  year = {2025},
  dimensions = {false},
}

EMSOFT’24

Efficient Discovery of Actual Causality Using Abstraction Refinement

Arshia Rafieioskouei, and Borzoo Bonakdarpour

IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, Nov 2024

Bib PDF Code Poster Slides Website

@article{10745858,
  author = {Rafieioskouei, Arshia and Bonakdarpour, Borzoo},
  journal = {IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems},
  title = {Efficient Discovery of Actual Causality Using Abstraction Refinement},
  year = {2024},
  month = nov,
  volume = {43},
  number = {11},
  pages = {4274-4285},
  url = {https://ieeexplore.ieee.org/document/10745858},
  dimensions = {false},
  doi = {10.1109/TCAD.2024.3448299}
}

CEJ

Ai-driven prediction of cardio-oncology biomarkers through protein corona analysis

Avirup Guha, Seyed Amirhossein Sadeghi, Harikrishnan Hyma Kunhiraman, and 14 more authors

Chemical Engineering Journal, Nov 2025

Code Website