I will give a talk at the UIUC Machine Learning Seminar. Thanks for the invitation and looking forward to connecting and engaging in discussions!

I will attend INFORMS 2026 in San Francisco, CA. Looking forward to connecting!

Congrats to Liuyuan Jiang on receiving the 2026 IEEE Signal Processing Society (SPS) Scholarship!

I will attend Asilomar 2026 at Pacific Grove, CA. Looking forward to meeting old and new friends there!

Our paper on uniform coverage of the Pareto front is accepted by NeurIPS 2026. Congrats to Liuyuan Jiang and Chentong Huang! We will present this work in Atlanta, GA, this December.
SURF: Steering the Scalarization Weight to Uniformly Traverse the Pareto Front
Preliminary slides Code Short description BibTeX

SURF studies how scalarization weights control the traversal speed along the Pareto front by analyzing the Pareto front geometry. Building on this, SURF proposes a geometry-guided weight sampling rule for uniform Pareto-front coverage. We apply the method to multi-objective RL and multi-objective LLM alignment.

@inproceedings{jiang2026surf,
  author    = {Jiang, Liuyuan and Huang, Chentong and Chen, Lisha},
  title     = {{SURF}: Steering the Scalarization Weight to Uniformly Traverse the {Pareto} Front},
  booktitle = {Advances in Neural Information Processing Systems},
  year      = {2026},
  url       = {https://arxiv.org/abs/2605.20619}
}

Happy to give a talk at OSU ECE department seminar on our recent works for preference-guided multi-objective learning. Thanks for the invitation and host!

I will present a tutorial at NeurIPS 2026 on Multi-Objective Optimization for Modern Machine Learning. More information will be available soon.

Grateful to receive the support from Tinker Research Grant and COE.

One paper accepted by Asilomar 2026. Congrats to Chentong Huang!

One paper accepted by ICASSP 2026 as oral. Congrats to Liuyuan Jiang!
BIRQ: Bi-level self-labeling random quantization for self-supervised speech recognition

Grateful to receive the support from Empire AI Beta!

A new preprint on bilevel optimization is out.
Efficient Penalty-Based Bilevel Methods: Improved Analysis, Novel Updates, and Flatness Condition

Two papers accepted by NeurIPS 2025. Congrats to Liuyuan Jiang and AFM Saif!
Beyond Value Functions: Single-loop Bilevel Optimization under Flatness Conditions

Objective Soups: Multilingual Multi-Task Modeling for Speech Processing

Our paper is accepted by ICML 2025 as spotlight.
Efficient First-Order Optimization on the Pareto Set for Multi-Objective Learning under Preference Guidance

I have successfully defended my thesis. Sincere thanks to my committee members for their guidance and support.

Our paper is accepted by IEEE Transactions on Audio, Speech and Language Processing.
Bilevel Joint Unsupervised and Supervised Training for Automatic Speech Recognition

Honored and grateful to receive the IEEE Signal Processing Society (SPS) scholarship.

Our paper is accepted by NeurIPS 2024.
FERERO: A Flexible Framework for Preference-Guided Multi-Objective Learning

I co-organize a special session on "Advances in Optimization for Machine Learning and Signal Processing" at IEEE SAM 2024 Workshop with Prof. Tianyi Chen and Prof. Kaiyi Ji.

Our paper is accepted by Interspeech 2024.
M2ASR: Multilingual Multi-task Automatic Speech Recognition via Multi-objective Optimization

Our paper is accepted by JMLR.
Three-Way Trade-Off in Multi-Objective Learning: Optimization, Generalization and Conflict-Avoidance

I present a tutorial at ICASSP 2024 together with Prof. Tianyi Chen and Dr. Xiaodong Cui.

I present a tutorial at AAAI 2024 together with Prof. Tianyi Chen and Prof. Zhuoran Yang.

I am invited to give a talk at the University of Rochester ECE department seminar.

Our paper is accepted by ICASSP 2024.
Variance Reduction Can Improve Trade-off in Multi-Objective Learning

Our NeurIPS23 paper is accepted by CPAL Spotlight Track.

I am honored to receive the IEEE Signal Processing Society (SPS) scholarship.

I am honored to receive the Rensselaer's Founders Award of Excellence.

Our paper is accepted by NeurIPS 2023. Check out the preprint:
Three-Way Trade-Off in Multi-Objective Learning: Optimization, Generalization and Conflict-Avoidance
I will present this work at:
- MoDL Meeting, New York City, NY     - EaGL Workshop, Rochester, NY
- Asilomar, Pacific Grove, CA                - NeurIPS, New Orleans, LA
Looking forward to meeting you there!

Our paper is accepted by ICASSP 2023
A Nested Ensemble Method to Bilevel Machine Learning

I am honored to be selected as Top Reviewer for AISTATS 2023.

Our monograph with Prof. Osvaldo Simeone's group is published in Foundations and Trends in Signal Processing
Learning with Limited Samples -- Meta-Learning and Applications to Communication Systems

Our paper is accepted by NeurIPS 2022
Understanding Benign Overfitting in Gradient-Based Meta Learning

Our paper is accepted by ICML 2022
Sharp-MAML: Sharpness-Aware Model-Agnostic Meta Learning

Our paper is accepted by AISTATS 2022
Is Bayesian Model-Agnostic Meta Learning Better than Model-Agnostic Meta Learning, Provably?

I am honored to receive the inaugural Belsky Award for Computational Sciences and Engineering from RPI.