Tue 6 Nov 2018 14:15 - 14:37 at Horizons 6-9F - Deep Learning Chair(s): David Rosenblum

With the rise of machine learning, there is a great deal of interest in treating programs as data to be fed to learning algorithms. However, programs do not start off in a form that is immediately amenable to most off-the-shelf learning techniques. Instead, it is necessary to transform the program to a suitable representation before a learning technique can be applied.

In this paper, we use abstractions of traces obtained from symbolic execution of a program as a representation for learning word embeddings. We trained a variety of word embeddings under hundreds of parameterizations, and evaluated each learned embedding on a suite of different tasks. In our evaluation, we obtain 93% top-1 accuracy on a benchmark consisting of over 19,000 API-usage analogies extracted from the Linux kernel. In addition, we show that embeddings learned from (mainly) semantic abstractions provide nearly triple the accuracy of those learned from (mainly) syntactic abstractions.

Tue 6 Nov

Displayed time zone: Guadalajara, Mexico City, Monterrey change

13:30 - 15:00
Deep LearningResearch Papers at Horizons 6-9F
Chair(s): David Rosenblum National University of Singapore
13:30
22m
Talk
Deep Learning Type Inference
Research Papers
Vincent J. Hellendoorn University of California at Davis, USA, Christian Bird Microsoft Research, Earl T. Barr , Miltiadis Allamanis Microsoft Research, Cambridge
13:52
22m
Talk
DeepSim: Deep Learning Code Functional Similarity
Research Papers
Gang Zhao , Jeff Huang Texas A&M University
14:15
22m
Talk
Code Vectors: Understanding Programs Through Embedded Abstracted Symbolic Traces
Research Papers
Jordan Henkel University of Wisconsin–Madison, Shuvendu K. Lahiri Microsoft Research, Ben Liblit University of Wisconsin–Madison, Thomas Reps University of Wisconsin - Madison and GrammaTech, Inc.
14:37
22m
Talk
MODE: Automated Neural Network Model Debugging via State Differential Analysis and Input Selection
Research Papers
Shiqing Ma Purdue University, USA, Yingqi Liu Purdue University, USA, Wen-Chuan Lee Purdue University, Xiangyu Zhang Purdue University, Ananth Grama Purdue University, USA