Yue Zhang

Yue Zhang

Research Scientist @ Meta | GenAI

I am a Research Scientist at Meta, where I develop memory and learning methods that enable LLM agents to retain, retrieve, update, and reason over experience across long-horizon interactions. Prior to Meta, I was a Machine Learning Engineer at Visa, and I earned my Ph.D. in Electrical Engineering and Computer Science from UC Merced, focusing on machine learning-augmented ubiquitous computing systems. I hold M.S. and B.S. degrees in Electronic Engineering from Tsinghua University. Please feel free to contact me via email (yzhang58 at ucmerced dot edu).

Research Interests

  • Agent Memory & Personalization
  • LLM Post-training & Reward Design
  • Retrieval-Grounded Generation & Tool Use
  • Agent Evaluation & Benchmarking

Education

  • Ph.D. in EECS

    UC Merced • 2024

  • M.S. in Electronic Engineering

    Tsinghua University • 2019

  • B.S. in Electronic Engineering

    Tsinghua University • 2016

Professional Experience

Research Scientist

Meta • Jun 2025 - Present

  • Leading research on memory systems that enable long-horizon reasoning and personalization in AI agents.
  • Advancing post-training for retrieval-grounded agents, improving query planning and grounded answer generation.
  • Designing agent evaluation benchmarks and reward signals for post-training.

Machine Learning Engineer

Visa • Aug 2024 - Jun 2025

  • Developed and onboarded machine learning-enabled real-time fraud detection and payment authorization models.
  • Optimized machine learning inference infrastructure for large-scale real-time service systems.

Ph.D. Researcher

UC Merced • Sep 2019 - Aug 2024

  • Developed a machine learning-augmented multimodal occupant sensing system.
  • Studied multimodal representation learning in time series data for cross-modal association.
  • Enhanced image generation from imbalanced datasets using contrastive learning.
  • Built a benchmark for causal reasoning over structured sensing data and evaluated frontier and open-weight LLMs on it.

Publications

Are Large Language Models Capable of Causal Reasoning for Sensing Data Analysis?

Zhizhang Hu, Yue Zhang, Ryan Rossi, Tong Yu, Sungchul Kim, Shijia Pan • The Workshop on Edge and Mobile Foundation Models, 2024

Poster: PrivaSee: Augmented Reality-Enabled Privacy Perception Visualization for Internet of Things

Yue Zhang*, Shangjie Du*, Jiqing Wen, Robert Likamwa, Shiwei Fang, Shijia Pan • The 22nd Annual International Conference on Mobile Systems, Applications and Services (MobiSys ’24), 2024

Best Poster Runner-up

Ioteeth: Intra-oral Teeth Sensing System for Dental Occlusal Diseases Recognition

Zhizhang Hu, Amirmohammad Radmehr, Yue Zhang, Shijia Pan, Phuc Nguyen • The ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies (UbiComp ’24), 2024

Poster: LEVO: LEGO® Bricks Enhanced Single-Point Vibration Sensing for Occupant Monitoring

Yue Zhang, Shikha Patel, Dong Yoon Lee, Paolo Celli, Amelie Bonde, Shijia Pan • The 21st ACM Conference on Embedded Networked Sensor Systems (SenSys ’23), 2023

Best Poster

CMA: Cross-Modal Association Between Wearable and Structural Vibration Signal Segments for Indoor Occupant Sensing

Yue Zhang, Zhizhang Hu, Uri Berger, Shijia Pan • The 22nd International Conference on Information Processing in Sensor Networks (IPSN ’23), 2023

VMA: Domain Variance- and Modality-Aware Model Transfer for Fine-Grained Occupant Activity Recognition

Zhizhang Hu, Yue Zhang, Tong Yu, Shijia Pan • 21st ACM/IEEE International Conference on Information Processing in Sensor Networks (IPSN ’22), 2022

AutoQual: Task-oriented Structural Vibration Sensing Quality Assessment Leveraging Co-located Mobile Sensing Context

Yue Zhang, Zhizhang Hu, Susu Xu, Shijia Pan • CCF Transactions on Pervasive Computing and Interaction, 2021

Patents

System and Method for Vibration-Based Medication Interaction Recognition

US 2025/0152091 A1 • Published 2025 • The Regents of the University of California

Pending

Personalized, Context Aware Input Output Systems and Methods

WO 2025/058720 A1 • Published 2025 • Futurewei Technologies

Pending

Cross-Modal Association Between Wearable and Structural Vibration Signal Segments for Indoor Occupant Sensing

US 2024/0364560 A1 • Published 2024 • The Regents of the University of California

Pending

Programmable Smart-Sensing Shelf Liner Apparatus for Autonomous Retail

US 2024/0138597 A1 • Published 2024 • The Regents of the University of California, AiFi Inc.

Pending

Indoor Power Line Positioning System and Method

CN 107942286 B • Granted 2020 • Tsinghua Shenzhen International Graduate School

Granted

Honors & Awards

Best Poster Runner-up Award

MobiSys 2024

Graduate Dean's Dissertation Fellowship

UC Merced 2024

Best Poster Award

SenSys 2023

Best Poster Runner-up Award

IPSN 2023

Best Demo Award

SenSys 2022

Best Poster Award

IPSN 2017, 2022

China National Scholarship

Tsinghua University 2019