• Publications
  • Influence
The Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networks
TLDR
In this paper, we show that there consistently exist sparse networks that train from the start and learn at least as fast as the original network while matching its test accuracy. Expand
  • 608
  • 103
  • PDF
Automatically patching errors in deployed software
We present ClearView, a system for automatically patching errors in deployed software. ClearView works on stripped Windows x86 binaries without any need for source code, debugging information, orExpand
  • 384
  • 30
  • PDF
Dynamic knobs for responsive power-aware computing
TLDR
We present PowerDial, a system for dynamically adapting application behavior to execute successfully in the face of load and power fluctuations. Expand
  • 300
  • 18
  • PDF
Using Datalog with Binary Decision Diagrams for Program Analysis
TLDR
This paper describes bddbddb, a BDD-Based Deductive DataBase, which implements the declarative language Datalog with stratified negation, totally-ordered finite domains and comparison operators. Expand
  • 193
  • 16
  • PDF
The Lottery Ticket Hypothesis: Training Pruned Neural Networks
TLDR
The lottery ticket hypothesis proposes that successful training depends on lucky random initialization of a smaller subcomponent of the network. Expand
  • 171
  • 14
Verifying quantitative reliability for programs that execute on unreliable hardware
TLDR
We present Rely, a language that enables developers to reason about the quantitative reliability of an application – namely, the probability that it produces the correct result when executed on unreliable hardware. Expand
  • 188
  • 13
  • PDF
Stabilizing the Lottery Ticket Hypothesis
TLDR
Pruning is a well-established technique for removing unnecessary structure from neural networks after training to improve the performance of inference. Expand
  • 57
  • 12
  • PDF
Chisel: reliability- and accuracy-aware optimization of approximate computational kernels
TLDR
We present Chisel, a system for reliability- and accuracy-aware optimization of approximate computational kernels that run on approximate hardware platforms. Expand
  • 144
  • 9
  • PDF
Context-sensitive program analysis as database queries
TLDR
This paper presents a new framework, based on the concept ofdeductive databases, for context-sensitive program analysis, in which all program information is stored as relations;data access and analyses are written as Datalog queries. Expand
  • 173
  • 8
  • PDF
Reflective program generation with patterns
TLDR
In this paper we examine a subclass of problems that can be addressed using a simpler mechanism than runtime reflection, which we call compile-time reflection. Expand
  • 41
  • 7
  • PDF