Sijia Liu

On the Convergence of A Class of Adam-Type Algorithms for Non-Convex\n Optimization

2018 · arXiv (Cornell University)

This paper studies a class of adaptive gradient based momentum algorithms\nthat update the search directions and learning rates simultaneously using past\ngradients. This class, which we refer to as the "Adam-type", includes the\npopular algorithms such as …

An ADMM-Based Universal Framework for Adversarial Attacks on Deep Neural Networks

2018

Deep neural networks (DNNs) are known vulnerable to adversarial attacks. That is, adversarial examples, obtained by adding delicately crafted distortions onto original legal inputs, can mislead a DNN to classify them as any target labels. …

Fast Training of Provably Robust Neural Networks by SingleProp

2021 · arXiv (Cornell University)

Recent works have developed several methods of defending neural networks against adversarial attacks with certified guarantees. However, these techniques can be computationally costly due to the use of certification during training. We develop a new …

TextGrad: Advancing Robustness Evaluation in NLP by Gradient-Driven Optimization

2022 · arXiv (Cornell University)

Robustness evaluation against adversarial examples has become increasingly important to unveil the trustworthiness of the prevailing deep models in natural language processing (NLP). However, in contrast to the computer vision domain where the first-order projected …

MERCY: Multiple Response Ranking Concurrently in Realistic Open-Domain Conversational Systems

2023

Automatic Evaluation (AE) and Response Selection (RS) models assign quality scores to various candidate responses and rank them in conversational setups. Prior response ranking research compares various models’ performance on synthetically generated test sets. In …