Alexander Wei
I'm Alex Wei, a fourth-year Computer Science PhD student at UC Berkeley advised by Nika Haghtalab, Michael I. Jordan, and Jacob Steinhardt. My research is centered around the intersections of machine learning, economics, and algorithm design. I am in particular excited about developing principles for learning and decision-making in complex environments.
Before Berkeley, I received my A.B. and S.M. in 2020 from Harvard, where I studied computer science, mathematics, and economics. There, I was fortunate to be advised by Jelani Nelson and Scott Kominers. The summer after college, I was an intern at Microsoft Research New England with Brendan Lucier and Nicole Immorlica. Last summer, I interned at FAIR with Anton Bakhtin and Noam Brown as a member of the Diplomacy project.
My work is supported by the NSF GRFP and a Meta Research PhD Fellowship.
Preprints
denotes alphabetical order
Jailbroken: How Does LLM Safety Training Fail?
arXiv, 2023
NeurIPS 2023 Oral Presentation (to appear)
@article{wei2023jailbroken,
title={Jailbroken: How Does LLM Safety Training Fail?},
author={Alexander Wei and Nika Haghtalab and Jacob Steinhardt},
journal={arXiv preprint arxiv:2307.02483},
year={2023}
}
Journal Papers
denotes alphabetical order
Learning Equilibria in Matching Markets from Bandit Feedback
Journal of the ACM, 2023
NeurIPS 2021 Spotlight Presentation
@article{jagadeesan2023learning,
title={Learning Equilibria in Matching Markets with Bandit Feedback},
author={Jagadeesan, Meena and Wei, Alexander and Wang, Yixin and Jordan, Michael I and Steinhardt, Jacob},
journal={Journal of the ACM},
volume={70},
number={3},
pages={1--46},
year={2023}
}
Human-Level Play in the Game of Diplomacy by Combining Language Models with Strategic Reasoning
Science, 2022
@article{fair2022diplomacy,
author = {Meta Fundamental AI Research Diplomacy Team (FAIR) and Anton Bakhtin and Noam Brown and Emily Dinan and Gabriele Farina and Colin Flaherty and Daniel Fried and Andrew Goff and Jonathan Gray and Hengyuan Hu and Athul Paul Jacob and Mojtaba Komeili and Karthik Konath and Minae Kwon and Adam Lerer and Mike Lewis and Alexander H. Miller and Sasha Mitts and Adithya Renduchintala and Stephen Roller and Dirk Rowe and Weiyan Shi and Joe Spisak and Alexander Wei and David Wu and Hugh Zhang and Markus Zijlstra},
title = {Human-level play in the game of \emph{Diplomacy} by combining language models with strategic reasoning},
journal = {Science},
volume = {378},
number = {6624},
pages = {1067--1074},
year = {2022},
}
Designing Approximately Optimal Search on Matching Platforms
Management Science, 2022
INFORMS Auctions & Market Design Rothkopf Junior Researcher Paper Prize, 3rd place
@article{immorlica2022designing,
author = {Immorlica, Nicole and Lucier, Brendan and Manshadi, Vahideh and Wei, Alexander},
title = {Designing Approximately Optimal Search on Matching Platforms},
journal = {Management Science},
volume = {69},
number = {8},
pages = {4609--4626},
year = {2022},
}
Optimal Las Vegas Approximate Near Neighbors in
ACM Transactions on Algorithms, 2022
SODA 2019 Best Student Paper
@article{wei2022optimal,
title={Optimal {Las Vegas} Approximate Near Neighbors in $\ell_p$},
author={Wei, Alexander},
journal={ACM Transactions on Algorithms},
volume={18},
number={1},
pages={1--27},
year={2022}
}
Conference Papers
denotes alphabetical order
TCT: Convexifying Federated Learning using Bootstrapped Neural Tangent Kernels
NeurIPS 2022
@article{yu2022tct,
title={{TCT}: Convexifying Federated Learning using Bootstrapped Neural Tangent Kernels},
author={Yu, Yaodong and Wei, Alexander and Karimireddy, Sai Praneeth and Ma, Yi and Jordan, Michael},
journal={Advances in Neural Information Processing Systems},
volume={35},
pages={30882--30897},
year={2022}
}
More Than a Toy: Random Matrix Models Predict How Real-World Neural Representations Generalize
ICML 2022
@inproceedings{wei2022more,
title={More Than a Toy: Random Matrix Models Predict How Real-World Neural Representations Generalize},
author={Wei, Alexander and Hu, Wei and Steinhardt, Jacob},
booktitle={Proceedings of the 39th International Conference on Machine Learning},
pages={23549--23588},
year={2022}
}
Predicting Out-of-Distribution Error with the Projection Norm
ICML 2022
@inproceedings{yu2022predicting,
title={Predicting Out-of-Distribution Error with the Projection Norm},
author={Yu, Yaodong and Yang, Zitong and Wei, Alexander and Ma, Yi and Steinhardt, Jacob},
booktitle={Proceedings of the 39th International Conference on Machine Learning},
pages={25721--25746},
year={2022}
}
Learning in Stackelberg Games with Non-myopic Agents
EC 2022
@inproceedings{haghtalab2022learning,
title={Learning in {Stackelberg} Games with Non-myopic Agents},
author={Haghtalab, Nika and Lykouris, Thodoris and Nietert, Sloan and Wei, Alexander},
booktitle={Proceedings of the 23rd ACM Conference on Economics and Computation},
pages={917--918},
year={2022}
}
Learning Equilibria in Matching Markets from Bandit Feedback
NeurIPS 2021
Spotlight Presentation
@article{jagadeesan2021learning,
title={Learning equilibria in matching markets from bandit feedback},
author={Jagadeesan, Meena and Wei, Alexander and Wang, Yixin and Jordan, Michael and Steinhardt, Jacob},
journal={Advances in Neural Information Processing Systems},
volume={34},
pages={3323--3335},
year={2021}
}
Designing Approximately Optimal Search on Matching Platforms
EC 2021
INFORMS Auctions & Market Design Rothkopf Junior Researcher Paper Prize, 3rd place
@inproceedings{immorlica2021designing,
title={Designing approximately optimal search on matching platforms},
author={Immorlica, Nicole and Lucier, Brendan and Manshadi, Vahideh and Wei, Alexander},
booktitle={Proceedings of the 22nd ACM Conference on Economics and Computation},
pages={632--633},
year={2021}
}
Optimal Robustness-Consistency Trade-Offs for Learning-Augmented Online Algorithms
NeurIPS 2020
@article{wei2020optimal,
title={Optimal robustness-consistency trade-offs for learning-augmented online algorithms},
author={Wei, Alexander and Zhang, Fred},
journal={Advances in Neural Information Processing Systems},
volume={33},
pages={8042--8053},
year={2020}
}
Better and Simpler Learning-Augmented Online Caching
APPROX 2020
@inproceedings{wei2020better,
title={Better and Simpler Learning-Augmented Online Caching},
author={Wei, Alexander},
booktitle={Approximation, Randomization, and Combinatorial Optimization. Algorithms and Techniques (APPROX/RANDOM 2020)},
pages={60:1--60:17},
year={2020}
}
Allocation for Social Good: Auditing Mechanisms for Utility Maximization
EC 2019
@inproceedings{lundy2019allocation,
title={Allocation for social good: auditing mechanisms for utility maximization},
author={Lundy, Taylor and Wei, Alexander and Fu, Hu and Kominers, Scott Duke and Leyton-Brown, Kevin},
booktitle={Proceedings of the 2019 ACM Conference on Economics and Computation},
pages={785--803},
year={2019}
}
Optimal Las Vegas Approximate Near Neighbors in
SODA 2019
Best Student Paper
Invited to the SODA 2019 special issue of ACM Transactions on Algorithms
@inproceedings{wei2019optimal,
title={Optimal {Las Vegas} Approximate Near Neighbors in $\ell_p$},
author={Wei, Alexander},
booktitle={Proceedings of the Thirtieth Annual ACM-SIAM Symposium on Discrete Algorithms},
pages={1794--1813},
year={2019}
}
Varying the Number of Signals in Matching Markets
WINE 2018
@inproceedings{jagadeesan2018varying,
title={Varying the number of signals in matching markets},
author={Jagadeesan, Meena and Wei, Alexander},
booktitle={International Conference on Web and Internet Economics},
pages={232--245},
year={2018}
}
Selected Awards
Meta Research PhD Fellowship (2022-2024)
NSF Graduate Research Fellowship (2020-2025)
Siebel Scholarship (2019-2020)
SODA Best Student Paper (2019)