Kuan Wang

Title: **Kuan Wang: Innovator in Neural Network Quantization**

Introduction

Kuan Wang is a prominent inventor based in Beijing, CN, noted for his significant contribution to the field of neural networks. He holds one patent that showcases his expertise and innovative approach in the technology sector, particularly in optimizing low-bit neural networks.

Latest Patents

Kuan Wang's latest patent is titled "Loss-error-aware quantization of a low-bit neural network". This patent encompasses methods, apparatus, systems, and articles of manufacture aimed at enhancing the quantization process of low-bit neural networks. The invention details an apparatus comprising a network weight partitioner that divides unquantized network weights of a primary network model into two distinct groups: one designated for quantization and the other for retraining. Additionally, the apparatus features a loss calculator for processing network weights to compute initial losses, a weight quantizer for generating low-bit network weights, and a weight updater that adjusts the second group of weights based on the loss differences. Finally, a network model deployer is included to implement the low-bit network model incorporating the revised weights.

Career Highlights

Kuan Wang is currently employed at Intel Corporation, where he employs his skills and knowledge to push the boundaries of innovation in technology. His work in the field of low-bit neural networks demonstrates his commitment to advancing adaptive technologies that have far-reaching implications in artificial intelligence and machine learning.

Collaborations

Throughout his career, Kuan has collaborated with talented coworkers such as Anbang Yao and Aojun Zhou. These partnerships are indicative of the collaborative spirit prevalent in research and development, contributing to the enhancement of technology through collective expertise.

Conclusion

Kuan Wang stands out as a key inventor in the field of neural networks, exemplifying innovation through his patented technology. His contributions not only reflect his individual talent but also highlight the importance of collaboration in the pursuit of technological advancements at Intel Corporation. As research continues to evolve, Kuan's work will likely play a crucial role in shaping the future of neural network applications.