Penghang Yin

Adversarial Defense via Data Dependent Activation Function and Total Variation Minimization

2018 · arXiv (Cornell University)

We improve the robustness of Deep Neural Net (DNN) to adversarial attacks by using an interpolating function as the output activation. This data-dependent activation remarkably improves both the generalization and robustness of DNN. In the …

Non-Ergodic Convergence Analysis of Heavy-Ball Algorithms

2019 · Proceedings of the AAAI Conference on Artificial Intelligence

In this paper, we revisit the convergence of the Heavy-ball method, and present improved convergence complexity results in the convex setting. We provide the first non-ergodic O(1/k) rate result of the Heavy-ball algorithm with constant …

Recurrence of Optimum for Training Weight and Activation Quantized Networks

2020 · arXiv (Cornell University)

Deep neural networks (DNNs) are quantized for efficient inference on resource-constrained platforms. However, training deep learning models with low-precision weights and activations involves a demanding optimization task, which calls for minimizing a stage-wise loss function …

COMQ: A Backpropagation-Free Algorithm for Post-Training Quantization

2025 · IEEE Access

Post-training quantization (PTQ) has emerged as a practical approach to compress large neural networks, making them highly efficient for deployment. However, effectively reducing these models to their low-bit counterparts without compromising the original accuracy remains …

DiaBlo: Diagonal Blocks Are Sufficient For Finetuning

2025 · arXiv (Cornell University)

Fine-tuning is a critical step for adapting large language models (LLMs) to domain-specific downstream tasks. To mitigate the substantial computational and memory costs of full-model fine-tuning, Parameter-Efficient Fine-Tuning (PEFT) methods have been proposed to update …