Papers I Read Notes and Summaries

Hamiltonian Neural Networks

Introduction

  • The paper proposes a very cool idea at the intersection of...


Extrapolating Beyond Suboptimal Demonstrations via Inverse Reinforcement Learning from Observations

Introduction

  • The paper proposes a new inverse RL (IRL) algorithm, called as...


Meta-Reinforcement Learning of Structured Exploration Strategies

Introduction

  • The paper looks at the problem of learning structured exploration policies...


Relational Reinforcement Learning

Introduction

  • Relational Reinforcement Learning (RRL) paradigm uses relational state (and action) space...


Good-Enough Compositional Data Augmentation

Introduction

  • The paper introduces a simple data augmentation protocol that provides a...


Multiple Model-Based Reinforcement Learning

  • The paper presents some general ideas and mechanisms for multiple model-based RL. Even...


Towards a natural benchmark for continual learning

Introduction

  • Continual Learning paradigm focuses on learning from a non-stationary stream of...


Meta-Learning Update Rules for Unsupervised Representation Learning

Introduction

  • Standard unsupervised learning aims to learn transferable features. The paper proposes...


GNN Explainer - A Tool for Post-hoc Explanation of Graph Neural Networks

Introduction

  • Graph Neural Network (GNN) is a family of powerful machine learning...


To Tune or Not to Tune? Adapting Pretrained Representations to Diverse Tasks