Papers I Read Notes and Summaries

HyperNetworks

Introduction

  • The paper explores HyperNetworks. The idea is to use one network...


Energy-based Models for Continual Learning

Introduction

  • The paper proposes to use Energy-based Models (EBMs) for Continual Learning.

    ...

GPipe - Easy Scaling with Micro-Batch Pipeline Parallelism

Introduction

  • The paper introduces GPipe, a pipeline parallelism library for scaling networks...


Compositional Explanations of Neurons

Introduction

  • The paper describes a method to explain/interpret the representations learned by...


Design patterns for container-based distributed systems

Introduction

  • The paper describes three design patterns for container-based distributed systems: single-container pattern,...

Cassandra - a decentralized structured storage system

Introduction

  • Cassandra is a distributed storage system that runs over cheap commodity servers...

CAP twelve years later - How the rules have changed

Introduction

  • The CAP theorem states that any system sharing data over the...


Consistency Tradeoffs in Modern Distributed Database System Design

Introduction

  • CAP theorem has been influential in the design decisions for distributed...


Exploring Simple Siamese Representation Learning

Introduction

  • The paper shows that Siamese networks can be used for unsupervised...


Data Management for Internet-Scale Single-Sign-On

Introduction

  • The paper describes the architecture of an erstwhile single-sign-on (SSO) service...