Controllability-aware unsupervised skill discovery

S Park, K Lee, Y Lee, P Abbeel - arXiv preprint arXiv:2302.05103, 2023 - arxiv.org
One of the key capabilities of intelligent agents is the ability to discover useful skills without
external supervision. However, the current unsupervised skill discovery methods are often …

Metra: Scalable unsupervised rl with metric-aware abstraction

S Park, O Rybkin, S Levine - arXiv preprint arXiv:2310.08887, 2023 - arxiv.org
Unsupervised pre-training strategies have proven to be highly effective in natural language
processing and computer vision. Likewise, unsupervised reinforcement learning (RL) holds …

PEAC: Unsupervised Pre-training for Cross-Embodiment Reinforcement Learning

C Ying, Z Hao, X Zhou, X Xu, H Su, X Zhang… - arXiv preprint arXiv …, 2024 - arxiv.org
Designing generalizable agents capable of adapting to diverse embodiments has achieved
significant attention in Reinforcement Learning (RL), which is critical for deploying RL …

Constrained Intrinsic Motivation for Reinforcement Learning

X Zheng, X Ma, C Shen, C Wang - arXiv preprint arXiv:2407.09247, 2024 - arxiv.org
This paper investigates two fundamental problems that arise when utilizing Intrinsic
Motivation (IM) for reinforcement learning in Reward-Free Pre-Training (RFPT) tasks and …

Augmenting Unsupervised Reinforcement Learning with Self-Reference

A Zhao, E Zhu, R Lu, M Lin, YJ Liu, G Huang - arXiv preprint arXiv …, 2023 - arxiv.org
Humans possess the ability to draw on past experiences explicitly when learning new tasks
and applying them accordingly. We believe this capacity for self-referencing is especially …

A Survey on Deep Learning-based Resource Allocation Schemes

D Kim, H Jung, IH Lee - 2023 14th International Conference on …, 2023 - ieeexplore.ieee.org
The growing number of complex and heterogeneous nodes and base station applications
has required a high computational complexity to handle wireless resources. To tackle this …

UNeC: Unsupervised Exploring In Controllable Space

X Xiong, L Meng, J Ruan, S Xu… - ICASSP 2024-2024 IEEE …, 2024 - ieeexplore.ieee.org
In unsupervised reinforcement learning, agents traverse a reward-free environment, aiming
for rapid generalisation to subsequent tasks. This strategy offers a compelling resolution to …

Surprise-Adaptive Intrinsic Motivation for Unsupervised Reinforcement Learning

A Hugessen, RC Castanyer, F Mohamed… - arXiv preprint arXiv …, 2024 - arxiv.org
Both entropy-minimizing and entropy-maximizing (curiosity) objectives for unsupervised
reinforcement learning (RL) have been shown to be effective in different environments …

CIM: Constrained Intrinsic Motivation for Reinforcement Learning

X Zheng, X Ma, C Shen, C Wang - openreview.net
This paper investigates two fundamental problems that arise when implementing intrinsic
motivation for reinforcement learning: 1) how to design a proper intrinsic objective for …

A Lightweight Siamese Network-Driven Unsupervised Reinforcement Learning

Z Ma, Y Huang - Available at SSRN 4706109 - papers.ssrn.com
Unsupervised reinforcement learning (URL) has been a promising paradigm in recent years
to design reinforcement learning (RL) agents that can be generalized to new tasks …