research Hardware Metric Reasoning using LLMs A benchmark for Verilog code metric reasoning using fine-tuned Large Language Models (LLMs) Low-Precision Training of Deep Neural Network An INT8-quantized training framework built upon the Forward-Forward algorithm, enabling efficient and memory-light deep learning training on resource-constrained edge devices Large-scale Approximate Computing Approximate logic synthesis using Boolean Matrix Factorization Dynamic Neural Network with Weight-Enabling Scheme A dynamic neural network framework that adaptively enables or disables weights during inference to balance accuracy and efficiency
