Get started
Author

Hui Zang

Observed name: Zang, Hui
Observed institution: Huawei Technologies (China)
OpenAlex ID: A5151780011
Pangu Pro MoE: Mixture of Grouped Experts for Efficient Sparsity
2025 · Tang, Yehui, Li, Xiaosong, et al. · arXiv (Cornell University) · 1
PDF
Trillion-Parameter MoE in a Box: Decoupling Memory Provisioning with High-Bandwidth Flash
2026 · Xia, Pengfei, Hao, Tuo, et al. · arXiv (Cornell University)
PDF
Enabling High-Bandwidth Flash for Generative Recommendation Serving with Write-Aware KV Cache Policy
2026 · Peng, Danni, Wu, Kai, et al. · arXiv (Cornell University)
PDF
Beyond Uniform Experts: Cost-Aware Expert Execution for Efficient Multi-Device MoE Inference
2026 · Zang, Hui, Xia, Pengfei, et al. · arXiv (Cornell University)
PDF
FlexNPU: Transparent NPU Virtualization for Dynamic LLM Prefill-Decode Co-location
2026 · Gu, Jiongjiong, Wang, Jianfeng, et al. · arXiv (Cornell University)
PDF
When KV Cache Reuse Fails in Multi-Agent Systems: Cross-Candidate Interaction is Crucial for LLM Judges
2026 · Association for Computational Linguistics 2026, Jia, Chujiajia, et al. · Underline Science Inc.
CONCAT: Consensus- and Confidence-Driven Ad Hoc Teaming for Efficient LLM-Based Multi-Agent Systems
2026 · Ma, Ziyang, Zhang, Dingyi, et al. · arXiv (Cornell University)
PDF
When KV Cache Reuse Fails in Multi-Agent Systems: Cross-Candidate Interaction is Crucial for LLM Judges
2026 · Sichu Liang, Zhenglin Wang, et al. · Proceedings of the conference - Association for Computational Linguistics. Meeting
PDF
Works count: 8
Citation count: 1
H-index: 1
I10-index:
Year
Topic
Type