Gotta Learn Fast: A New Benchmark for Generalization in RL

Abstract

In this report, we present a new reinforcement learning (RL) benchmark based on the Sonic the Hedgehog (TM) video game franchise. This benchmark is intended to measure the performance of transfer learning and few-shot learning algorithms in the RL domain. We also present and evaluate some baseline algorithms on the new benchmark.

References in corpus (6)

• Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks
• Matching Networks for One Shot Learning
• A Simple Neural Attentive Meta-Learner
• PathNet: Evolution Channels Gradient Descent in Super Neural Networks
• Actor-Mimic: Deep Multitask and Transfer Reinforcement Learning
• DARLA: Improving Zero-Shot Transfer in Reinforcement Learning

Cited by in corpus (58)

• Unity: A General Platform for Intelligent Agents
• Meta-World: A Benchmark and Evaluation for Multi-Task and Meta Reinforcement Learning
• Quantifying Generalization in Reinforcement Learning
• Leveraging Procedural Generation to Benchmark Reinforcement Learning
• A Survey of Deep Reinforcement Learning in Video Games
• Rapid trial-and-error learning with simulation supports flexible tool use and physical reasoning
• Assessing Generalization in Deep Reinforcement Learning
• Generalization and Regularization in DQN
• Improving Generalization in Meta Reinforcement Learning using Learned Objectives
• Meta reinforcement learning as task inference
• Automatic Data Augmentation for Generalization in Deep Reinforcement Learning
• Neural MMO: A Massively Multiagent Game Environment for Training and Evaluating Intelligent Agents
• The NetHack Learning Environment
• The MineRL 2019 Competition on Sample Efficient Reinforcement Learning using Human Priors
• RIDE: Rewarding Impact-Driven Exploration for Procedurally-Generated Environments
• Off-Policy Evaluation via Off-Policy Classification
• See, Hear, Explore: Curiosity via Audio-Visual Association
• Observational Overfitting in Reinforcement Learning
• Meta-learning curiosity algorithms
• Using a thousand optimization tasks to learn hyperparameter search strategies
• Invariant Policy Optimization: Towards Stronger Generalization in Reinforcement Learning
• Extending Environments To Measure Self-Reflection In Reinforcement Learning
• Task-Agnostic Dynamics Priors for Deep Reinforcement Learning
• Why Generalization in RL is Difficult: Epistemic POMDPs and Implicit Partial Observability
• Improving the Generalization of End-to-End Driving through Procedural Generation
• Arena: A General Evaluation Platform and Building Toolkit for Multi-Agent Intelligence
• Fever Basketball: A Complex, Flexible, and Asynchronized Sports Game Environment for Multi-agent Reinforcement Learning
• Measuring Visual Generalization in Continuous Control from Pixels
• Reinforcement Learning Generalization with Surprise Minimization
• Retrospective Analysis of the 2019 MineRL Competition on Sample Efficient Reinforcement Learning
• rl_reach: Reproducible Reinforcement Learning Experiments for Robotic Reaching Tasks
• The MAGICAL Benchmark for Robust Imitation
• Toybox: A Suite of Environments for Experimental Evaluation of Deep Reinforcement Learning
• Using Unity to Help Solve Intelligence
• When Is Generalizable Reinforcement Learning Tractable?
• Transfer of Fully Convolutional Policy-Value Networks Between Games and Game Variants
• Rank the Episodes: A Simple Approach for Exploration in Procedurally-Generated Environments
• Systematic Evaluation of Causal Discovery in Visual Model Based Reinforcement Learning
• TeachMyAgent: a Benchmark for Automatic Curriculum Learning in Deep RL
• Decoupling Value and Policy for Generalization in Reinforcement Learning
• Procedural Generalization by Planning with Self-Supervised World Models
• Model Based Planning with Energy Based Models
• Scenic4RL: Programmatic Modeling and Generation of Reinforcement Learning Environments
• Arena: a toolkit for Multi-Agent Reinforcement Learning
• Tuning Mixed Input Hyperparameters on the Fly for Efficient Population Based AutoRL
• The Principle of Unchanged Optimality in Reinforcement Learning Generalization
• Sparse Attention Guided Dynamic Value Estimation for Single-Task Multi-Scene Reinforcement Learning
• Generalization to Novel Objects using Prior Relational Knowledge
• Dynamic Value Estimation for Single-Task Multi-Scene Reinforcement Learning
• Online Safety Assurance for Deep Reinforcement Learning
• Meta Arcade: A Configurable Environment Suite for Meta-Learning
• Towards robust and domain agnostic reinforcement learning competitions
• Object-Oriented Dynamics Learning through Multi-Level Abstraction
• Neural MMO v1.3: A Massively Multiagent Game Environment for Training and Evaluating Neural Networks
• Supervise Thyself: Examining Self-Supervised Representations in Interactive Environments
• Extrapolation in Gridworld Markov-Decision Processes
• Enhanced Scene Specificity with Sparse Dynamic Value Estimation
• Lifetime policy reuse and the importance of task capacity