Go-Explore: a New Approach for Hard-Exploration Problems
- Adrien Ecoffet, Joost Huizinga, J. Lehman, Kenneth O. Stanley, J. Clune
- Computer ScienceArXiv
- 30 January 2019
A new algorithm called Go-Explore, which exploits the following principles to remember previously visited states, solve simulated environments through any available means, and robustify via imitation learning, which results in a dramatic performance improvement on hard-exploration problems.
First return then explore
- Adrien Ecoffet, Joost Huizinga, J. Lehman, Kenneth O. Stanley, J. Clune
- Computer ScienceThe Naturalist
- 27 April 2020
The substantial performance gains from Go-Explore suggest that the simple principles of remembering states, returning to them, and exploring from them are a powerful and general approach to exploration-an insight that may prove critical to the creation of truly intelligent learning agents.
Scaling MAP-Elites to deep neuroevolution
- Cédric Colas, Joost Huizinga, Vashisht Madhavan, J. Clune
- Computer ScienceAnnual Conference on Genetic and Evolutionary…
- 3 March 2020
A new hybrid algorithm called MAP-Elites with Evolution Strategies (ME-ES) is designed and evaluated for post-damage recovery in a difficult high-dimensional control task where traditional ME fails, and it is shown that ME-ES performs efficient exploration, on par with state-of-the-art exploration algorithms in high- dimensional control tasks with strongly deceptive rewards.
Video PreTraining (VPT): Learning to Act by Watching Unlabeled Online Videos
- Bowen Baker, Ilge Akkaya, J. Clune
- Computer ScienceArXiv
- 23 June 2022
This work extends the internet-scale pretraining paradigm to sequential decision domains through semi-supervised imitation learning wherein agents learn to act by watching online unlabeled videos, and is the first to report computer agents that can craft diamond tools.
The Evolutionary Origins of Hierarchy
- H. Mengistu, Joost Huizinga, Jean-Baptiste Mouret, J. Clune
- Biology, Computer SciencePLoS Comput. Biol.
- 23 May 2015
The results suggest that the same force–the cost of connections–promotes the evolution of both hierarchy and modularity, and that these properties are important drivers of network performance and adaptability.
Evolving neural networks that are both modular and regular: HyperNEAT plus the connection cost technique
- Joost Huizinga, J. Clune, Jean-Baptiste Mouret
- Computer ScienceAnnual Conference on Genetic and Evolutionary…
- 12 July 2014
It is shown that adding the connection cost technique to Hyper NEAT produces neural networks that are significantly more modular, regular, and higher performing than HyperNEAT without a connection cost, even when compared to a variant of HyperNEat that was specifically designed to encourage modularity.
The Emergence of Canalization and Evolvability in an Open-Ended, Interactive Evolutionary System
- Joost Huizinga, Kenneth O. Stanley, J. Clune
- BiologyArtificial Life
- 17 April 2017
It is revealed that genomes entrench certain dimensions of variation that were frequently explored during their evolutionary history, and it is shown that these organizational properties correlate with increased fitness.
Evolving Multimodal Robot Behavior via Many Stepping Stones with the Combinatorial Multiobjective Evolutionary Algorithm
- Joost Huizinga, J. Clune
- Computer ScienceEvolutionary Computation
- 9 July 2018
A thorough introduction and investigation of the Combinatorial Multiobjective Evolutionary Algorithm (CMOEA), which allows all combinations of subtasks to be explored simultaneously, and shows that adding a linear combination over all objectives can improve the ability of the control algorithms to solve these multimodal problems.
Does Aligning Phenotypic and Genotypic Modularity Improve the Evolution of Neural Networks?
- Joost Huizinga, Jean-Baptiste Mouret, J. Clune
- BiologyAnnual Conference on Genetic and Evolutionary…
- 20 July 2016
Results suggest encouraging modularity in both the genotype and phenotype as an important step towards solving large-scale multi-modal problems, but also indicate that more research is required before the authors can evolve structurally organized networks to solve tasks that require multiple, different neural modules.
Exploration Based Language Learning for Text-Based Games
- Andrea Madotto, Mahdi Namazifar, G. Tur
- Computer ScienceInternational Joint Conference on Artificial…
- 24 January 2020
This work presents an exploration and imitation-learning-based agent capable of state-of-the-art performance in playing text-based computer games and shows that the learned policy can generalize better than existing solutions to unseen games without using any restriction on the action space.
...
...