Short Biography:
I am a research scientist at the Laboratory for Information and Decision Systems (LIDS). Prior to that, I was a postdoctoral research fellow at MIT from 2014 to 2017. I received my BS from Sharif University of Technology in Tehran, Iran, and my PhD in electrical and computer engineering from Concordia University in Montreal, Canada, in 2013. I have been the recipient of several awards, including two gold medals in the International Mathematical Olympiad (IMO), Concordia University Doctoral Prize in Engineering and Computer Science, Governor General of Canada Academic Gold Medal, and NSERC Postdoctoral Fellowship.
Teaching:
I have over ten years experience in teaching the art of problem solving to students preparing to participate in National and International Mathematical Olympiad. Courses I have taught at MIT:
See my Teaching page for more details.
Research:
My recent research is primarily focused on two fronts:
- Social learning: I strive to develop a profound understanding of learning and decision making under uncertainty at both microscopic and systemic levels in large population of strategic agents under computational, communication, and cognitive constraints. This requires incorporating tools and techniques from game theory, finite/infinite dimensional optimization, control theory, modern probability theory, and statistics. Some of my recent work in this area:
- Transformers and self-attention networks: My aim is to make a better understanding of the inner working of transformers and attention mechanisms by analyzing them through the lens of control theory and dynamical systems. Recent related work:
I am also interested in applied research in revenue management and pricing:
Please see my Research and Publications to learn more about my research.
Contact Information:
Laboratory for Information and Decision Systems
Institute for Data, Systems, and Society
Massachusetts Institute of Technology
32-D572 Vassar Street,
Cambridge, MA 02139
Phone: (215) 919-3234
Email: ajorlou (at) mit (dot) edu
I am a research scientist at the Laboratory for Information and Decision Systems (LIDS). Prior to that, I was a postdoctoral research fellow at MIT from 2014 to 2017. I received my BS from Sharif University of Technology in Tehran, Iran, and my PhD in electrical and computer engineering from Concordia University in Montreal, Canada, in 2013. I have been the recipient of several awards, including two gold medals in the International Mathematical Olympiad (IMO), Concordia University Doctoral Prize in Engineering and Computer Science, Governor General of Canada Academic Gold Medal, and NSERC Postdoctoral Fellowship.
Teaching:
I have over ten years experience in teaching the art of problem solving to students preparing to participate in National and International Mathematical Olympiad. Courses I have taught at MIT:
- Nonlinear Optimization (6.252): Course syllabus . Link to lecture notes.
- Introduction to Network Models (1.022): Course material available on MIT OpenCourseWare. See 1.022 final project (Fall 2020) on adaptive modeling and simulation of COVID-19 spread in python.
See my Teaching page for more details.
Research:
My recent research is primarily focused on two fronts:
- Social learning: I strive to develop a profound understanding of learning and decision making under uncertainty at both microscopic and systemic levels in large population of strategic agents under computational, communication, and cognitive constraints. This requires incorporating tools and techniques from game theory, finite/infinite dimensional optimization, control theory, modern probability theory, and statistics. Some of my recent work in this area:
- Estimating true beliefs from declared opinions, with J. Tang, A. Adler, and A. Jadbabaie, IEEE Transactions on Automatic Control, conditionally accepted.
- Stochastic opinion dynamics under social pressure in arbitrary networks, with J. Tang, A. Adler, and A. Jadbabaie, IEEE Transactions on Automatic Control, under review.
- Belief samples are all you need for social learning, with M. JafariNodeh and A. Jadbabaie, 63rd IEEE Conference on Decision and Control, under review.
- A game-theoretic model of misinformation spread on social networks, with C. Hsu and A. Jadbabaie, Games and Economic Behavior, under review.
- Subscription networks, verification, and media bias, with C. Hsu, M. Yildiz, and A. Jadbabaie, Review of Economic Studies, under review.
- Transformers and self-attention networks: My aim is to make a better understanding of the inner working of transformers and attention mechanisms by analyzing them through the lens of control theory and dynamical systems. Recent related work:
- On the role of attention masks and LayerNorm in self-attention dynamics, with X. Wu, Y. Wang, S. Jegelka, and A. Jadbabaie, Advances in Neural Information Processing Systems (NeurIPS), 2024, under review.
- Demystifying oversmoothing in attention-based graph neural networks, with X. Wu, Z. Wu, and A. Jadbabaie, in Advances in Neural Information Processing Systems (NeurIPS), 2023 (spotlight paper).
I am also interested in applied research in revenue management and pricing:
- Sales-Based rebate design, with A. Jadbabaie, Management Science, vol. 69, pp. 5983–6000, 2023.
- Dynamic pricing in social networks: The word of mouth effect, with A. Jadbabaie and A. Kakhbod, Management Science, vol. 64, pp. 971-979, 2018.
Please see my Research and Publications to learn more about my research.
Contact Information:
Laboratory for Information and Decision Systems
Institute for Data, Systems, and Society
Massachusetts Institute of Technology
32-D572 Vassar Street,
Cambridge, MA 02139
Phone: (215) 919-3234
Email: ajorlou (at) mit (dot) edu