Huihong Shi

3 papers · 2023–2024 · 2 venues · across top CS/AI venues

Authors are matched by name, so researchers who share a name ("Yang Liu", "Wei Wang") appear as a single profile and their paper counts, venues and achievements are combined. How identity is handled.

Achievements

Jump to papers ↓

🌍 Conference Polyglot · Bronze 🌉 Interdisciplinary Bridge · Bronze 🗺️ Taxonomy Completionist (10) 🐣 Hot Topic Early Bird

Venues

NeurIPS (2) ICML (1)

Top co-authors

Haoran You (2) Yipin Guo (2) Yonggan Fu (1) Amir Yazdanbakhsh (1) Yingyan (Celine) Lin (1) Xiaofan Zhang (1) Wei Zhou (1) Zheng Wang (1) Yingyan Celine Lin (1) Zhongzhi Yu (1)

Research topics

Artificial Intelligence (2) Deep Learning (2) Machine Learning (2) Natural Language Processing (2) Application Areas (2) Core AI (2) Optimization & Theory · Machine Learning (2) Optimization & Theory · Deep Learning (2) Resources & Methods (2) Architectures (1)

Keywords

large language model (2) model compression (2) attention mechanism (1) efficient computing (1) mixture of expert (1) weight quantization (1) latency reduction (1) inference acceleration (1) multiplication-free model (1) hardware acceleration (1) attention sink (1) attention calibration (1) bit allocation (1) accuracy improvement (1) multiplication reduction (1) shift-add reparameterization (1) shift-and-add reparameterization (1) inference (1) input-adaptive (1) vision transformer (1)

Papers

ShiftAddLLM: Accelerating Pretrained LLMs via Post-Training Multiplication-Less Reparameterization NeurIPS 2024 Unveiling and Harnessing Hidden Attention Sinks: Enhancing Large Language Models without Training through Attention Calibration ICML 2024 ShiftAddViT: Mixture of Multiplication Primitives Towards Efficient Vision Transformer NeurIPS 2023