I’m a final-year Ph.D. candidate in the Centre for Intelligent Machines (CIM) at McGill University, and a core contributor at 2077AI-Foundation. Prior to my Ph.D., I received my Master’s degree in Electrical Computer Engineering (ECE) and Computational Science and Engineering (CSE) from Georgia Institute of Technology in 2022. I earned my Bachelor’s degree in Electrical Engineering from Northeastern University.

My research explores the intersection of Machine Learning, Large Language Models, and Autonomous Systems, with a particular focus on evaluation, robustness, and real-world reliability.

I design structured benchmarks, agent environments, and reward models to evaluate Vision-Language (VLM) and Vision-Language-Action (VLA) models in multiple complex, long-horizon scenarios, including safety-critical domains such as autonomous driving. My work studies how agents reason, perceive, and act under uncertainty, distribution shifts, and adversarial conditions. More broadly, I am interested in bridging static model benchmarks and operational agent evaluation, enabling models in settings that more closely resemble real-world deployment across diverse and complex environments.

📬 I’m always open to collaborations—please feel free to reach out!

📢 I am actively looking for industrial internships and full-time opportunities. I am happy to engage in discussions regarding potential opportunities!

Selected Research Works

* indicates equal contribution

🚗 VLM & VLA for Autonomous Driving

AgentThink: Tool-Augmented Reasoning in VLMs for Autonomous Driving
K. Qian*, S. Jiang*, Y. Zhong*, Z. Luo, Z. Huang, et al.
EMNLP 2025

A Survey on Vision–Language–Action Models for Autonomous Driving
S. Jiang*, Z. Huang*, K. Qian*, Z. Luo, T. Zhu, et al.
ICCV-W, 2025

FASIONAD+:Enhanced Safety in Autonomous Driving with Adaptive Feedback
Z. Luo*, S. Jiang*, K. Qian*, Z. Huang, J. Miao, et al.
Under review, 2025

Communication-Aware Reinforcement Learning for Cooperative Adaptive Cruise Control
S. Jiang, S. Choi, L. Sun
TRB Annual Meeting (Oral), 2024

⚖️ Foundation Models: Evaluation, Alignment & Safety

EditReward: A Human-Aligned Reward Model for Instruction-Guided Image Editing
K. Wu*, S. Jiang*, M. Ku, P. Nie, M. Liu, W. Chen
Under review, 2025

Adversarial Vulnerabilities in Large Language Models for Time Series Forecasting
F. Liu*, S. Jiang*, L. Miranda-Moreno, S. Choi, L. Sun
AISTATS 2025

VeriWeb: Verifiable Long-Chain Web Benchmark for Agentic Information-Seeking
2077AI Team
Under review, 2025

Temporally Sparse Attack for Fooling Large Language Models in Time Series Forecasting
F. Liu, S. Jiang
TrustworthyLLM @ ICLR 2025

News

November 2025
One paper accepted as an oral presentation at Bridge Program on AAAI 2026.

October 2025
Attended COLM 2025 and ICCV 2025! Abaka AI is honored to be one of the largest sponsor at ICCV 2025!

August 2025
One paper AgentThink got accepted at EMNLP 2025!

August 2025
I’m thrilled to join the 2077AI-Foundation—contributing to an open-source ecosystem that powers next-gen foundation-model datasets for everyone! 🌟 Let’s build the future of AI together!

August 2025
Excited to join Abaka AI as a Research Scientist in Mountain View, California! 🚀 I’ll be working on next-generation auto-labeling frameworks for multimodal foundation models and autonomous driving data — open to collaboration and discussion!

July 2025
Our survey paper “A Survey on Vision-Language-Action Models for Autonomous Driving” is accepted at Foundation Models for Autonomous Driving on ICCV 2025.

June 2025
We released the first comprehensive survey on Vision–Language–Action (VLA) models for autonomous driving, in collaboration with Tsinghua University, Xiaomi, and the University of Wisconsin–Madison.
📄 Paper Link
🌟 The companion repo Awesome-VLA4AD received over 100 stars within the first three days!
📰 Coverage by several tech media outlets: [AutoDriving Heart] [Embodied Evolution] [Zhihu]

May 2025
Our collaborative work with Tsinghua and Xiaomi, AgentThink: A Unified Framework for Tool-Augmented VLM Reasoning in Autonomous Driving! This is the first framework that integrates tool-augmented VLM reasoning in driving tasks — thank you Xiaomi for the support!
📄 Paper Link

April 2025
I was invited to contribute to the Humanity’s Last Exam benchmark — a milestone initiative for evaluating AGI safety. Grateful to be part of this important effort!

Feb 2025
Our work Temporally Sparse Attack for Fooling Large Language Models in Time Series Forecasting was accepted to the ICLR 2025 Trustworthy LLM Workshop!
📄 Paper Link

Awards

• McGill Engineering Doctoral Award (MEDA), 2021–2024
• TISED Doctoral Recruitment Award (DRA), McGill University, 2021
• Outstanding Graduate of Liaoning Province, 2019
• Most Influential Graduate, Northeastern University, 2019
• 1st Class Academic Scholarship, Northeastern University, 2018
• National 1st Prize, China Undergraduate Mathematical Contest in Modeling, 2017
• 1nd Class Academic Scholarship, Northeastern University, 2016

Academic Service

Conference Reviewer:

• NeurIPS (Neural Information Processing Systems)
• ICLR (International Conference on Learning Representations)
• AISTATS (International Conference on Artificial Intelligence and Statistics)
• CVPR (Conference on Computer Vision and Pattern Recognition)
• ICCV (International Conference on Computer Vision)
• COLM (Conference on Language Modeling)
• EMNLP (Conference on Empirical Methods in Natural Language Processing)
• AAAI (Association for the Advancement of Artificial Intelligence)
• IROS (IEEE/RSJ International Conference on Intelligent Robots and Systems)
• ICRA (IEEE International Conference on Robotics and Automation)
• ITSC (IEEE International Conference on Intelligent Transportation Systems)

Journal Reviewer:

• IEEE Robotics and Automation Letters (RA-L)
• Transportation Research Part C: Emerging Technologies (TRC)
• IEEE Transactions on Intelligent Transportation Systems (T-ITS)