Exploration in deep reinforcement learning: A survey

P Ladosz, L Weng, M Kim, H Oh - Information Fusion, 2022‏ - Elsevier
This paper reviews exploration techniques in deep reinforcement learning. Exploration
techniques are of primary importance when solving sparse reward problems. In sparse …

Continual learning for robotics: Definition, framework, learning strategies, opportunities and challenges

T Lesort, V Lomonaco, A Stoian, D Maltoni, D Filliat… - Information fusion, 2020‏ - Elsevier
Continual learning (CL) is a particular machine learning paradigm where the data
distribution and learning objective change through time, or where all the training data and …

Guiding pretraining in reinforcement learning with large language models

Y Du, O Watkins, Z Wang, C Colas… - International …, 2023‏ - proceedings.mlr.press
Reinforcement learning algorithms typically struggle in the absence of a dense, well-shaped
reward function. Intrinsically motivated exploration methods address this limitation by …

Off-policy deep reinforcement learning without exploration

S Fujimoto, D Meger, D Precup - … conference on machine …, 2019‏ - proceedings.mlr.press
Many practical applications of reinforcement learning constrain agents to learn from a fixed
batch of data which has already been gathered, without offering further possibility for data …

Go-explore: a new approach for hard-exploration problems

A Ecoffet, J Huizinga, J Lehman, KO Stanley… - arxiv preprint arxiv …, 2019‏ - arxiv.org
A grand challenge in reinforcement learning is intelligent exploration, especially when
rewards are sparse or deceptive. Two Atari games serve as benchmarks for such hard …

Cooperative exploration for multi-agent deep reinforcement learning

IJ Liu, U Jain, RA Yeh… - … conference on machine …, 2021‏ - proceedings.mlr.press
Exploration is critical for good results in deep reinforcement learning and has attracted much
attention. However, existing multi-agent deep reinforcement learning algorithms still use …

Skew-fit: State-covering self-supervised reinforcement learning

VH Pong, M Dalal, S Lin, A Nair, S Bahl… - arxiv preprint arxiv …, 2019‏ - arxiv.org
Autonomous agents that must exhibit flexible and broad capabilities will need to be
equipped with large repertoires of skills. Defining each skill with a manually-designed …

Reconciling novelty and complexity through a rational analysis of curiosity.

R Dubey, TL Griffiths - Psychological Review, 2020‏ - psycnet.apa.org
Curiosity is considered to be the essence of science and an integral component of cognition.
What prompts curiosity in a learner? Previous theoretical accounts of curiosity remain …

Evolution-guided policy gradient in reinforcement learning

S Khadka, K Tumer - Advances in Neural Information …, 2018‏ - proceedings.neurips.cc
Abstract Deep Reinforcement Learning (DRL) algorithms have been successfully applied to
a range of challenging control tasks. However, these methods typically suffer from three core …

Revisiting rainbow: Promoting more insightful and inclusive deep reinforcement learning research

JSO Ceron, PS Castro - International Conference on …, 2021‏ - proceedings.mlr.press
Since the introduction of DQN, a vast majority of reinforcement learning research has
focused on reinforcement learning with deep neural networks as function approximators …