CURL: Contrastive Unsupervised Representations for Reinforcement Learning
Contrastive Learning has been an established method in NLP and Image classification. The authors show that with relatively minor adjustments, CL can be used to augment and improve RL dramatically.
Paper: https://arxiv.org/abs/2004.04136
Code: https://github.com/MishaLaskin/curl
Abstract:
We present CURL: Contrastive Unsupervised Representations for Reinforcement Learning. CURL extracts high-level features from raw pixels using contrastive learning and performs off-policy control on top of the extracted features. CURL outperforms prior pixel-based methods, both model-based and model-free, on complex tasks in the DeepMind Control Suite and Atari Games showing 2.8x and 1.6x performance gains respectively at the 100K interaction steps benchmark. On the DeepMind Control Suite, CURL is the first image-based algorithm to nearly match the sample-efficiency and performance of methods that use state-based features.
Authors: Aravind Srinivas, Michael Laskin, Pieter Abbeel
Links:
YouTube: https://www.youtube.com/c/yannickilcher
Twitter: https://twitter.com/ykilcher
BitChute: https://www.bitchute.com/channel/yannic-kilcher
Minds: https://www.minds.com/ykilcher
Видео CURL: Contrastive Unsupervised Representations for Reinforcement Learning канала Yannic Kilcher
Paper: https://arxiv.org/abs/2004.04136
Code: https://github.com/MishaLaskin/curl
Abstract:
We present CURL: Contrastive Unsupervised Representations for Reinforcement Learning. CURL extracts high-level features from raw pixels using contrastive learning and performs off-policy control on top of the extracted features. CURL outperforms prior pixel-based methods, both model-based and model-free, on complex tasks in the DeepMind Control Suite and Atari Games showing 2.8x and 1.6x performance gains respectively at the 100K interaction steps benchmark. On the DeepMind Control Suite, CURL is the first image-based algorithm to nearly match the sample-efficiency and performance of methods that use state-based features.
Authors: Aravind Srinivas, Michael Laskin, Pieter Abbeel
Links:
YouTube: https://www.youtube.com/c/yannickilcher
Twitter: https://twitter.com/ykilcher
BitChute: https://www.bitchute.com/channel/yannic-kilcher
Minds: https://www.minds.com/ykilcher
Видео CURL: Contrastive Unsupervised Representations for Reinforcement Learning канала Yannic Kilcher
Показать
Комментарии отсутствуют
Информация о видео
Другие видео канала
Supervised Contrastive LearningSupervised and Unsupervised Learning In Machine Learning | Machine Learning Tutorial | SimplilearnReinforcement Learning - "DDPG" explainedDeepMind Made A Superhuman AI For 57 Atari Games! 🕹Reconciling modern machine learning and the bias-variance trade-offMomentum Contrastive LearningFixMatch: Simplifying Semi-Supervised Learning with Consistency and ConfidenceWhat Is Self-Supervised Learning? | AI with AlexApps are dead... what's the next big thing?Lecture 7 Self-Supervised Learning -- UC Berkeley Spring 2020 - CS294-158 Deep Unsupervised LearningCURL: Contrastive Unsupervised Representations for Reinforcement LearningICML 2020 Oral Talk: Planning to Explore via Self-Supervised World Models[ML News] GPT-3 learns to edit | Google Pathways | Make-A-Scene | CLIP meets GamePhysics | DouBlindMomentum Contrast for Unsupervised Visual Representation LearningHow Reinforcement Learning can be Applied to Quantitative Finance w/ Dr. Tom StarkeTeachable Machine 1: Image ClassificationSelf-supervised denoising using blind-spot convolutional networksMIT 6.S091: Introduction to Deep Reinforcement Learning (Deep RL)Exploiting Latent Structure and Bisimulation Metrics for Better Generalization