Jie Fu (付杰) is a happy and funny visiting scholar at Hong Kong University of Science and Technology (HKUST), chasing his human-friendly big AI dream.

He was a postdoctoral fellow (funded by Microsoft Research Montreal) supervised by Yoshua Bengio at University of Montreal, Quebec AI Institute (Mila). He was an IVADO postdoctoral fellow supervised by Chris Pal at Polytechnique Montreal, Quebec AI Institute (Mila). He worked as a researcher (PI) at BAAI (智源人工智能研究院). He obtained his PhD from National University of Singapore under the supervision of Tat-Seng Chua. He received ICLR 2021 Outstanding Paper Award.

He is currently working towards safe, scalable system-2, mixed-modal interactive LLMs that are capable of observing, acting, and receiving feedback iteratively from external entities. More concretely:

  • Towards Safe and Scalable System-2 Language Models: Human-AI Alignment, Reinforcement learning, Meta learning, Lifelong learning, Reasoning, Modular neural architectures, etc
  • Applications: Multi-modal learning (e.g., NLP, CV), Multi-modal embodied intelligence, AI for science, etc

Selected Publications

I’m (still…) training myself (slowly…)

My full publications can be found in Google Scholar.
When Do Graph Neural Networks Help with Node Classification: Investigating the Homophily Principle on Node Distinguishability
NeurIPS 2023
MARBLE: Music Audio Representation Benchmark for Universal Evaluation
NeurIPS 2023
Running Ahead of Evolution - AI based Simulation for Predicting Future High-risk SARS-CoV-2 Variants
IJHPCA, ACM Gordon Bell COVID Finalist 2022
MERT: Acoustic Music Understanding Model with Large-Scale Self-supervised Training
arXiv
Think Before You Act: Decision Transformers with Internal Working Memory
arXiv
Interactive Natural Language Processing
arXiv
MUDiff: Unified Diffusion for Complete Molecule Generation
arXiv
Chinese Open Instruction Generalist: A Preliminary Release
arXiv
Learning Multi-Objective Curricula for Robotic Policy Learning
CoRL 2022
Biological Sequence Design with GFlowNets
ICML 2022
CoCon: A Self-Supervised Approach for Controlled Text Generation
ICLR 2021
Rikinet: Reading Wikipedia Pages for Natural Question Answering
ACL 2020
Interactive Machine Comprehension with Information Seeking Agents
ACL 2020