Название: Programming Languages and Systems: 19th Asian Symposium, APLAS 2021, Chicago, IL, USA, October 17–18, 2021, Proceedings
Автор: Hakjoo Oh (Ed.)
Издательство: Springer
Год: 2021
Страниц: 328
Язык: английский
Формат: pdf (true)
Размер: 14.2 MB

This book constitutes the proceedings of the 19th Asian Symposium on Programming Languages and Systems, APLAS 2021, held in Chicago, USA, in October 2021. The 17 papers presented in this volume were carefully reviewed and selected from 43 submissions. They were organized in topical sections named: analysis and synthesis, compilation and transformation, language, and verification.

Software applications and technologies are built on top of foundational systems such as compilers, databases, and theorem provers. Such foundations form the trusted computing base, and fundamentally impact software quality and security. Thus, it is a critical challenge to solidify and advance them. This book highlights general, effective techniques, and extensive, impactful efforts on finding hundreds of critical issues in widely-used compilers, database management systems, and SMT solvers. It focuses on the high-level principles and core techniques, their significant practical successes, and future opportunities and challenges.

As neural networks are trained to be deeper and larger, the scalability of neural network analyzer is urgently required. The main technical insight of our method is modularly analyzing neural networks by segmenting a network into blocks and conduct the analysis for each block. In particular, we propose the network block summarization technique to capture the behaviors within a network block using a block summary and leverage the summary to speed up the analysis process. We instantiate our method in the context of a CPU-version of the state-of-the-art analyzer DeepPoly and name our system as Bounded-Block Poly (BBPoly). We evaluate BBPoly extensively on various experiment settings. The experimental result indicates that our method yields comparable precision as DeepPoly but runs faster and requires less computational resources. Especially, BBPoly can analyze really large neural networks like SkipNet or ResNet that contain up to one million neurons in less than around 1 hour per input image, while DeepPoly needs to spend even 40 hours to analyze one image.