Scholar
Jierun Chen
Google Scholar ID: 8rPHNOsAAAAJ
HKUST
Multi-modal Models
Large Language Models
Efficient AI
Follow
Homepage
↗
Google Scholar
↗
Citations & Impact
All-time
Citations
2,354
H-index
7
i10-index
6
Publications
14
Co-authors
14
list available
Contact
Email
jierunchen@gmail.com
CV
Open ↗
GitHub
Open ↗
LinkedIn
Open ↗
Publications
6 items
FastCode: Fast and Cost-Efficient Code Understanding and Reasoning
2026
Cited
0
From Verifiable Dot to Reward Chain: Harnessing Verifiable Reference-based Rewards for Reinforcement Learning of Open-ended Generation
2026
Cited
0
SWE-Lego: Pushing the Limits of Supervised Fine-tuning for Software Issue Resolving
arXiv.org · 2026
Cited
5
InSight-o3: Empowering Multimodal Foundation Models with Generalized Visual Search
2025
Cited
0
Efficient Reasoning for Large Reasoning Language Models via Certainty-Guided Reflection Suppression
2025
Cited
0
The Synergy Dilemma of Long-CoT SFT and RL: Investigating Post-Training Techniques for Reasoning VLMs
2025
Cited
0
Resume (English only)
Academic Achievements
CVPR 2025: Proposed SnapGen—the first mobile-deployable text-to-image model (379M) generating 1024x1024 images in 1.4s with GenEval score of 0.66
CVPR Workshop BEAM 2025: Cleaned standard REC benchmarks and introduced Ref-L4, a new benchmark for referring expression comprehension in the LMM era
NeurIPS 2024: Introduced AsCAN—a hybrid network with asymmetric convolution-attention blocks for efficient recognition and generation
ICDE 2024: Proposed TASFAR, a target-agnostic source-free domain adaptation method for regression tasks
CVPR 2023: Developed PConv (Partial Convolution) and FasterNet, a latency-efficient architecture family
ICDE 2023: Built a semi-supervised geofencing system using ambient RF signals for applications like elderly care and pandemic control
Preprint 2023: Proposed StableKD—a stable and data-efficient knowledge distillation framework
Pacific Graphics 2023: Proposed CP-NeRF for cross-scene novel view synthesis with a single unified NeRF
ICDCS 2023: Designed FIS-ONE, a floor identification system requiring only one labeled RF sample
Co-authors
14 total
Co-author 1
Weipeng Zhuo
Beijing Normal-Hong Kong Baptist University
Chul-Ho Lee
Computer Science, Texas State University
Song Wen
Rutgers University
Hao He
Student, HKUST
Shiu-hong Kao
HKUST
Co-author 7
Tianlang He
The Hong Kong University of Science and Technology
×
Welcome back
Sign in to Agora
Welcome back! Please sign in to continue.
Email address
Password
Forgot password?
Continue
Do not have an account?
Sign up