Ph.D. Max Planck Institute for Intelligent Systems
About Me
I am a Research Group Leader in the Social Foundations of Computation Department at the Max Planck Institute for Intelligent Systems in Tübingen, Germany, hosted by Professor Moritz Hardt.
I study AI evaluation and algorithmic decision-making systems in which people, models, and institutions interact strategically. My work aims to understand how incentives and decision rules shape behavior and outcomes in these systems, and to design mechanisms that remain robust under strategic adaptation. Currently, I am particularly interested in these questions in LLM evaluation ecosystems.
I completed my Ph.D. from the Computer Science and Engineering Department at the University of California, Santa Cruz, where I was fortunate to be advised by Professor Yang Liu. I was supported by the Chancellor’s Dissertation-Year Fellowship. I was a student researcher at Google Brain hosted by Ehsan Amid and Rohan Anil in the summer of 2022.
I received a MS in Statistics at Stanford University in 2019, where I worked with Professor Johan Ugander on node classification problem using semi-supervised learning.
Before that, I received my BS in Energy and Resources Engineering and BA in Economics from Peking University in 2016.
[July 2026] I will give two talks on my recent paper about incentives in LLM leaderboards: one at Machine Learning for Science in Tübingen and another at WALE in Limnos, Greece!
Publications
(* equal contribution)
Leaderboard Incentives: Model Rankings under Strategic Post-Training Yatong Chen, Guanhua Zhang, and Moritz Hardt. ICML 2026. [arXiv]
ROC-n-reroll: How verifier imperfection affects test-time scaling
Florian E Dorner, Yatong Chen*, André F. Cruz*, and Fanny Yang. ICLR 2026. [arXiv]
Strategic Hypothesis Testing Yatong Chen*, Safwan Hossian*, and Yiling Chen. NeurIPS 2025. Spotlight, top 3%. [arXiv]
To Give or Not to Give? The Impacts of Strategically Withheld Recourse Yatong Chen, Andrew Estornell, Yevgeniy Vorobeychik and Yang Liu. AISTATS 2025. (Preliminary version at NeurIPS 2023 Workshop on Algorithmic Fairness through the Lens of Time). [arXiv]
Performative Prediction with Bandit Feedback: Learning through Reparameterization Yatong Chen, Wei Tang, Chien-Ju Ho and Yang Liu. ICML 2024. [ArXiv]
Learning to Incentivize Improvements from Strategic Agents Yatong Chen, Jialu Wang and Yang Liu. TMLR. (Preliminary version at ICML 2021 Workshop on Algorithmic Recourse, Best Paper Award). [paper][code]
Model Transferability with Responsive Decision Subjects Yatong Chen, Zeyu Tang, Kun Zhang and Yang Liu. ICML 2023. (Preliminary version at ICML 2022 Workshop on Adversarial Machine Learning Frontiers, Best Paper Award). [paper][talk slides][code]
Incentivizing Recourse through Auditing in Strategic Classification
Andrew Estornell, Yatong Chen, Sammy Das, Yang Liu, Yevgeniy Vorobeychik . IJCAI 2023. [paper]
Tier Balancing: Towards Dynamic Fairness over Underlying Causal Factors
Zeyu Tang, Yatong Chen, Yang Liu and Kun Zhang. ICLR 2023. (Preliminary version at NeurIPS 2022 Workshop on Algorthmic Fairness Through the Lens of Causality and Privacy). [paper]
Fair Transferability Subject to Distribution Shift Yatong Chen*, Reilly Raab*, Jialu Wang and Yang Liu. Neurips 2022. (Preliminary version at NeurIPS 2021 Workshop on Algorithmic Fairness through the Lens of Causality and Robustness). [arXiv]
Metric-Fair Classifier Derandomization
Jimmy Wu, Yatong Chen and Yang Liu. ICML 2022. Spotlight Presentation. [paper][talk slides]
Decoupled Smoothing on Graphs
Alex Chin, Yatong Chen, Kristen M. Altenburger and Johan Ugander. WWW 2019. Oral Presentation, top 5%. [paper][code][talk slides]
Workshop Papers
Fast Implicit Constrained Optimization of Non-decomposable Objectives for Deep Networks Yatong Chen, Abhishek Kumar, Yang Liu, Ehsan Amid. Has it Trained Yet? NeurIPS 2022 Workshop. [paper]
Fishy: Layerwise Fisher Approximation for Higher-order Neural Network Optimization
Abel Peirson*, Ehsan Amid*, Yatong Chen, Vladimir Feinberg, Manfred K Warmuth, Rohan Anil. Has it Trained Yet? NeurIPS 2022 Workshop. [paper]
Decoupled Smoothing In Probabilistic Soft Logic Yatong Chen*, Bryan Tor*, Eriq Augustine and Lise Getoor. 15th International Workshop on Mining and Learning with Graphs. [paper][talk slides]
Distinctions
Chancellor’s Dissertation-Year Fellowship (only one recipient in the School of Engineering), 2023, UC Santa Cruz
Best Paper Award, ICML Workshop on New Frontiers in Adversarial Machine Learning (AdvML), 2022
Finalist, Google PhD Fellowship Program, 2022
Best Paper Award, ICML Workshop on Algorithmic Recourse, 2021
Baskin Engineering Fellowship for Anti-Racism Research (FARR), 2021, UC Santa Cruz
Regents Fellowship, 2019, UC Santa Cruz
Selective Talks
Workshop on Algorithms for Learning and Economics (WALE), Limnos, Greece.
Invited talk on Why LLM Leaderboards Create Bad Incentives and How to Fix Them.
July 2026.
[link]
Machine Learning in Science Conference, Tübingen, Germany.
Invited talk on Why LLM Leaderboards Create Bad Incentives and How to Fix Them.
July 2026.
ETH AI Center Academic Talk Series (AICATS), Zürich, Switzerland.
Invited talk on Why LLM Leaderboards Create Bad Incentives and How to Fix Them.
March 2026.
Tübingen Women in Machine Learning (TWiML), Tübingen, Germany.
Invited talk on Why LLM Leaderboards Create Bad Incentives and How to Fix Them.
March 2026.
[link]
Friday Talks at the Tübingen AI Center, Tübingen, Germany.
Talk on Strategic Hypothesis Testing.
November 2025.
University of Pennsylvania Theory Seminar, Philadelphia, USA.
Invited talk on Metric-Fair Classifier Derandomization.
April 12, 2024.
International Conference on Machine Learning (ICML), Baltimore, USA.
Talk on Metric-Fair Classifier Derandomization.
July 19, 2022.
[link]
Workshop on New Frontiers in Adversarial Machine Learning (AdvML Frontiers),
ICML, Baltimore, USA.
Talk on Model Transferability With Responsive Decision Subjects.
July 22, 2022.
[link]
Workshop on Algorithmic Recourse, ICML, virtual.
Spotlight talk on Learning Classifiers that Incentivize Constructive Adaptation.
July 24, 2021.
[link]
Institute for Foundations of Data Science (IFDS) Ethics Seminar.
Invited talk on Learning Classifiers that Incentivize Constructive Adaptation.
May 2021.
Workshop on Consequential Decision Making in Dynamic Environments,
NeurIPS, virtual.
Talk on Learning Classifiers that Incentivize Constructive Adaptation.
December 12, 2020.
[link]
UCSC Theory Reading Group, Santa Cruz, USA.
Invited talk on Decoupled Smoothing on Graphs.
October 2019.
The Web Conference, San Francisco, USA.
Talk on Decoupled Smoothing on Graphs.
May 17, 2019.
[link]
