Sitemap

A list of all the posts and pages found on the site. For you robots out there is an XML version available for digesting as well.

Pages

Posts

Multi-Layer Perceptron from Scratch

less than 1 minute read

Published:

This is an implementation of multi-layer perceptron and its training with backpropagation from scratch using Numpy.

Reinforcement Learning Study Note

less than 1 minute read

Published:

There are already a number of excellent tutorials and lectures on Reinforcement Learning (RL) available, but I often find that these do not provide enough detail or explanation of the formulas behind them. This may be because a lot of the concepts are assumed to be obvious or straightforward.

portfolio

publications

A Unified Neural Coherence Model

Published in Journal 1, 2009

This paper is about the number 1. The number 2 is left for future work.

Recommended citation: Your Name, You. (2009). "Paper Title Number 1." Journal 1. 1(1).

talks

A Unified Neural Coherence Model

Published:

Recently, neural approaches to coherence modeling have achieved state-of-the-art results in several evaluation tasks. However, we show that most of these models often fail on harder tasks with more realistic application scenarios. In particular, the existing models underperform on tasks that require the model to be sensitive to local contexts such as candidate ranking in conversational dialogue and in machine translation. In this paper, we propose a unified coherence model that incorporates sentence grammar, inter-sentence coherence relations, and global coherence patterns into a common neural framework. With extensive experiments on local and global discrimination tasks, we demonstrate that our proposed model outperforms existing models by a good margin, and establish a new state-of-the-art.
   

GradMask: Gradient-Guided Token Masking for Textual Adversarial Example Detection

Published:

We present GradMask, a simple adversarial example detection scheme for natural language processing (NLP) models. It uses gradient signals to detect adversarially perturbed tokens in an input sequence and occludes such tokens by a masking process. GradMask provides several advantages over existing methods including improved detection performance and an interpretation of its decision with a only moderate computational cost. Its approximated inference cost is no more than a single forward- and back-propagation through the target model without requiring any additional detection module. Extensive evaluation on widely adopted NLP benchmark datasets demonstrates the efficiency and effectiveness of GradMask.    

teaching

DeepNLP

Graduate course, University 1, Department, 2014

This is a description of a teaching experience. You can use markdown like any other post.

Teaching experience 2

Workshop, University 1, Department, 2015

This is a description of a teaching experience. You can use markdown like any other post.