I am a Ph.D. Candidate (2022.08-) in Computing and Information Sciences at Rochester Institute of Technology (RIT), advised by Prof. Weijie Zhao. I previously interned at ByteDance (AI Agents) and Amazon (Agents & Foundation Models). My research interests span GenAI, Agents, LLMs, AI Security, and Scalable & Trustworthy Machine Learning.
Prior to RIT, I worked as a Research Assistant at the Institute of Computer Vision (ICV), Shenzhen University, supervised by Prof. Linlin Shen (Honorary Professor at the University of Nottingham, UK). I received my B.S. degree in Computer Science and Technology in July 2021. During my undergraduate studies, I led a National Innovation Project funded by the Ministry of Education of China and received several ACM-ICPC Medals.
News
Sep 2026
Two papers are accepted by AACL-IJCNLP 2026: VTBench (Main) and UniGuardian (Findings)! 🎉🎉🎉
Aug 2026
Invited as a reviewer for ICLR 2027.
Aug 2026
Invited as a reviewer for IEEE Transactions on Circuits and Systems for Video Technology (TCSVT).
Jun 2026
Invited as a reviewer for AAAI 2027.
May 2026
We released MUSE-Autoskill, a self-evolving agent framework that enables agents to create, remember, manage, and evaluate reusable skills for more capable long-horizon problem solving.
May 2026
I will be attending CVPR 2026 from June 2 to June 8 to present our paper DMin. Looking forward to meeting you!
Mar 2026
Invited as a reviewer for NeurIPS 2026. (Declined, in protest of the sanctioned institutions policy; science should have no borders; as a reviewer, my role is to serve science, not politics).
Feb 2026
One paper is accepted by CVPR 2026 (Findings)! 🎉🎉🎉
Feb 2026
Invited as a reviewer for KDD 2026.
Jan 2026
One paper is accepted by ICLR 2026! 🎉🎉🎉
Jan 2026
Invited as a reviewer for ICML 2026.
Dec 2025
I will be joining ByteDance (TikTok) as a Research Intern in San Jose, CA in the spring of 2026, focusing on general open-world agents. 🎉🎉🎉
Oct 2025
I will be attending ICDM 2025 to deliver a tutorial titled "Behavior-Aware Data Valuation for LLMs at Scale" on Nov. 13 in Washington, D.C. I look forward to meeting you there! 🤗🤗🤗 [Slides]
Sep 2025
Invited as a reviewer for ICLR 2026.
Aug 2025
Invited as a reviewer for AAAI 2026.
May 2025
We released VTBench, a comprehensive benchmark evaluating over 20 visual tokenizers. [Paper], [Codebase], [Dataset], [Demo] are available.
May 2025
I will be joining Amazon as an Applied Scientist Intern in Bellevue, WA this summer, focusing on LLM agents. 🎉🎉🎉
Apr 2025
Apr 2025
Invited as a reviewer for NeurIPS 2025.
Apr 2025
One paper is accepted by SIGIR 2025! 🎉🎉🎉
Mar 2025
Invited as a reviewer for ACL ARR 2025.
Mar 2025
Invited as a reviewer for ICML 2025.
Jan 2025
Invited as a reviewer for IEEE Transactions on Dependable and Secure Computing.
Jan 2025
One paper is accepted by NAACL 2025! 🎉🎉🎉
Aug 2024
Invited as a reviewer for ICLR 2025.
Jul 2024
Invited as a reviewer for ACL ARR 2024.
Jul 2024
Invited as a reviewer for KDD 2025.
May 2024
Our paper (Token-wise Influential Training Data Retrieval) is accepted by ACL 2024!
Apr 2024
I will be joining Amazon as an Applied Scientist Intern in Boston this summer, focusing on LLM and multimodal systems. 🎉🎉🎉
Feb 2024
We released an easy-to-run implementation for finetuning LLMs. [Github Repo]
Feb 2024
Invited as a reviewer for KDD 2024.
Oct 2023
Invited as a reviewer for NeurIPS 2023 GLFrontiers Workshop.
Aug 2023
Received the KDD’23 Student Travel Award. Thanks to KDD!
May 2023
Sep 2022
Our paper (Activation Template Matching Loss) is accepted by FG 2023!
Publications
ArXiv 2026
MUSE-Autoskill: Self-Evolving Agents via Skill Creation, Memory, Management, and Evaluation
AACL-IJCNLP 2026
VTBench: Evaluating Visual Tokenizers for Autoregressive Image Generation
AACL-IJCNLP 2026 (Findings)
UniGuardian: A Unified Defense for Detecting Prompt Injection, Backdoor Attacks and Adversarial Attacks in Large Language Models
CVPR 2026 (Findings)
DMin: Scalable Training Data Influence Estimation for Diffusion Models
ArXiv 2025
Agent-Omni: Test-Time Multimodal Reasoning via Model Coordination for Understanding Anything
ArXiv 2025
Online Gradient Boosting Decision Tree: In-Place Updates for Efficient Adding/Deleting Data
ArXiv 2025
RapidUn: Influence-Driven Parameter Reweighting for Efficient Large Language Model Unlearning
NAACL 2025
ALinFiK: Learning to Approximate Linearized Future Influence Kernel for Scalable Third-Parity LLM Data Valuation
ACL 2024
Token-wise Influential Training Data Retrieval for Large Language Models
- Huawei Lin, Peng Li, Jie Song, Fuxin Jiang, Tieying Zhang, MUSE-Autoskill: Self-Evolving Agents via Skill Creation, Memory, Management, and Evaluation. arXiv preprint, arXiv:2605.27366, 2026.
- Huawei Lin, Tong Geng, Zhaozhuo Xu, Weijie Zhao, VTBench: Evaluating Visual Tokenizers for Autoregressive Image Generation. AACL-IJCNLP 2026.
- Huawei Lin, Yingjie Lao, Tong Geng, Tan Yu, Weijie Zhao, UniGuardian: A Unified Defense for Detecting Prompt Injection, Backdoor Attacks and Adversarial Attacks in Large Language Models. AACL-IJCNLP (Findings) 2026.
- Huawei Lin, Yingjie Lao, Weijie Zhao, DMin: Scalable Training Data Influence Estimation for Diffusion Models. CVPR (Findings) 2026.
- Ziwen Liu, Huawei Lin, Yide Ran, Denghui Zhang, Jianwen Xie, Chuan Li, Weijie Zhao, Zhaozhuo Xu, Randomized Antipodal Search Done Right for Data Pareto Improvement of LLM Unlearning. ICLR 2026.
- Huawei Lin, Yunzhi Shi, Tong Geng, Weijie Zhao, Wei Wang, Ravender Pal Singh, Agent-Omni: Test-Time Multimodal Reasoning via Model Coordination for Understanding Anything. arXiv preprint, arXiv:2511.02834, 2025.
- Huawei Lin, Jun Woo Chung, Yingjie Lao, Weijie Zhao, Online Gradient Boosting Decision Tree: In-Place Updates for Efficient Adding/Deleting Data. arXiv preprint, arXiv:2502.01634, 2025.
- Guoshenghui Zhao, Huawei Lin, Weijie Zhao, RapidUn: Influence-Driven Parameter Reweighting for Efficient Large Language Model Unlearning. arXiv preprint, arXiv:2512.04457, 2025.
- Jun Woo Chung, Huawei Lin, Weijie Zhao, Locality-Sensitive Indexing for Graph-Based Approximate Nearest Neighbor Search. SIGIR 2025.
- Yanzhou Pan, Huawei Lin, Yide Ran, Jiamin Chen, Xiaodong Yu, Weijie Zhao, Denghui Zhang, Zhaozhuo Xu, ALinFiK: Learning to Approximate Linearized Future Influence Kernel for Scalable Third-Parity LLM Data Valuation. NAACL 2025.
- Huawei Lin, Jikai Long, Zhaozhuo Xu, Weijie Zhao, Token-wise Influential Training Data Retrieval for Large Language Models. ACL 2024.
- Huawei Lin, Jun Woo Chung, Yingjie Lao, Weijie Zhao, Machine Unlearning in Gradient Boosting Decision Trees. KDD 2023.
- Huawei Lin, Haozhe Liu, Qiufu Li, Linlin Shen, Activation Template Matching Loss for Explainable Face Recognition. FG 2023.
Tech Talks
Jun 2026
"MUSE-Autoskill: Self-Evolving Agents via Skill Creation, Memory, Management, and Evaluation", ByteDance, San Jose, CA. [Slides]
Nov 2025
"Foundation of Data Valuation & Influence Estimation", ICDM, Washington, DC. [Slides]
May 2025
Apr 2025
"Influence Estimation for Large Language Models and Diffusion Models", RIT, Rochester, NY. [Slides]
May 2024
"Toward Explainable Large Language Models via Influence Estimation", Boston, MA. [Poster]
Aug 2023
Nov 2022
Oct 2021
"Toward Explainable Face Recognition", Shenzhen University, Shenzhen, China. [Slides]
Aug 2021
"Trust in Black-Box Models: Interpretability & Explainability for Deep Learning", Shenzhen University, Shenzhen, China. [Slides]
Experience
ByteDance - AI Infra
Jan 2026 - Jun 2026
Research Intern
San Jose, CA
Research Topic: MUSE-Autoskill: Self-Evolving Agents via Skill Creation, Memory, Management, and Evaluation
Advisor: Peng Li, Jie Song, Fuxin Jiang
Advisor: Peng Li, Jie Song, Fuxin Jiang
- Designed and implemented MUSE-AutoSkill Agent, an end-to-end framework for creating, retrieving, executing, and refining reusable skills.
- Built persistent skill infrastructure and skill-level memory to support skill management, reuse, and continual improvement across tasks.
- Integrated unit tests, execution feedback, and iterative correction into the skill lifecycle to improve agent reliability and adaptability.
Amazon
Applied Scientist Intern
Bellevue, WA
Research Topic: Unlocking Multi-modal Reasoning through Agent-Based Coordination
Advisor: Yunzhi Shi
Advisor: Yunzhi Shi
- Present a novel agent-based omni agent that coordinates existing foundation models to reason over text, images, video, and audio.
- Design a flexible agent system that interprets user intent and delegates subtasks.
Amazon
Applied Scientist Intern
Boston, MA
Research Topic: Auto Prompt: Unsupervised Self-improving Inference for LLMs
Advisor: Rajath Kumar, Raphael Petegrosso
Advisor: Rajath Kumar, Raphael Petegrosso
- Proposed an unsupervised self-improving framework for LLMs inference that enhances generation quality.
- Implemented methods to detect potential hallucinations by certainty score.
Shenzhen University (Institute of Computer Vision)
Research Assistant Intern
Shenzhen, China
Advisor: Linlin Shen
- Responsible for interpretability and explainability on deep learning (Biometrics).
- Proposed Explainable Channel Loss (ECLoss) for explainable face recognition.
ByteDance (AI Lab)
Oct 2020 - Jul 2021
Software Engineer Intern
Beijing, China
- Proposed Stream Audio Understanding Chain method for real-time processing.
- Optimized pipeline throughput by 65% using multiprocessing.
Honors
2023
KDD Student Travel Award
2020
Zhou Lian Scholarship (Top 0.005%)
2020
IAPR/IEEE WSB, Student Travel Grant
19-20
Technology Innovation Scholarship (x2)
18-21
ACM-ICPC Medals (Silver x1, Bronze x5)
Reviewer
2027
AAAI, ICLR
2026
NeurIPS, ICML, ICLR, KDD, AAAI
2025
NeurIPS, ICML, ICLR, KDD, ACL Rolling
2024
KDD, ACL Rolling
Journals
IEEE TDSC, IEEE TCSVT, IEEE Access
Beyond the Lab
My Road Trips & Ski Map
Miles driven. Mountains skied.
The colored lines trace my cross-country drives across America; the blue markers show ski resorts I have visited. Hover over a marker or resort name to explore.
12,600miles driven
10ski resorts
I-94 · Jun
I-90 · Sep
I-80 · Dec
Route 66 + Gulf Coast · Jul
Ski resort
Tahoe close-up Five nearby resorts
Markers show resort base areas, not ski-terrain boundaries. Palisades: Olympic Valley; Heavenly: California Lodge. Geography: Natural Earth.
My cross-country drivesSelect a trip to trace its route
Ski resorts I've visitedHover over a resort to preview it on the map