Scholar
Peijie Dong
Google Scholar ID: TqS6s4gAAAAJ
Ph.D. Candidate in HKUST(GZ)
Citations & Impact
Citations
838
H-index
17
i10-index
25
Publications
20
Co-authors
8
list available
Publications
Resume (English only)
Academic Achievements
- Publications: - ChunkKV: Semantic-Preserving KV Cache Compression for Efficient Long-Context LLM Inference (NeurIPS 2025) - Perovskite-LLM: Knowledge-Enhanced Large Language Models for Perovskite Solar Cell Research (EMNLP 2025 findings) - Smooth Reading: Bridging the Gap of Recurrent LLM to Self-Attention LLM on Long-Context Tasks (arxiv, 2025.08) - Intern-S1: A Scientific Multimodal Foundation Model (arxiv, 2025.08) - Can Compressed LLMs Truly Act? An Empirical Evaluation of Agentic Capabilities in LLM Compression (ICML25) - SpInfer: Leveraging Low-Level Sparsity for Efficient Large Language Model Inference on GPUs (EuroSys 2025, Best Paper) - STBLLM: Breaking the 1-Bit Barrier with Structured Binary LLMs (ICLR25) - The Lottery LLM Hypothesis (ICLR25 Blogpost Oral) - ParZC: Parametric Zero-Cost Proxies for Efficient NAS (AAA25 Oral) - FuseFL: One-Shot Federated Learning through the Lens of Causality with Progressive Layer Fusion (NeurIPS 2024 Spotlight) - Discovering Sparsity Allocation for Layer-wise Pruning of Large Language Models (NeurIPS 2024)
- Awards: - Excellent Research Prize (2024 DSA Excellent Research Award) - Invited as an Area Chair in NeurIPS 2025
- Other Achievements: - Gave a talk at PDL on 'Introduction to LLM Compression and Beyond'
Research Experience
- Projects include exploring techniques for model compression (pruning, quantization, knowledge distillation), optimizing large language models, and developing methods for automated machine learning.
Education
- Degree: PhD; School: The Hong Kong University of Science and Technology (Guangzhou); Field: Data Science and Analysis; Advisors: Prof. Xiaowen Chu, Prof. Junxian He.
Background
- Research Interests: model compression, efficient large language models, and machine learning systems. Background: PhD candidate in Data Science and Analysis at The Hong Kong University of Science and Technology (Guangzhou), supervised by Prof. Xiaowen Chu and Prof. Junxian He.
Miscellany
- Personal interests and hobbies not mentioned
