English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Pre-training a diverse set of neural network controllers in simulation hasenabled robots to adapt online to damage in robot locomotion tasks. However,finding diverse, high-performing controllers requires expensive networktraining and extensive tuning of a large number of hyperparameters. On theother hand, Covariance Matrix Adaptation MAP-Annealing (CMA-MAE), an evolutionstrategies (ES)-based quality diversity algorithm, does not have theselimitations and has achieved state-of-the-art performance on standard QDbenchmarks. However, CMA-MAE cannot scale to modern neural network controllersdue to its quadratic complexity. We leverage efficient approximation methods inES to propose three new CMA-MAE variants that scale to high dimensions. Ourexperiments show that the variants outperform ES-based baselines in benchmarkrobotic locomotion tasks, while being comparable with or exceedingstate-of-the-art deep reinforcement learning-based quality diversityalgorithms.

Training Diverse High-Dimensional Controllers by Scaling Covariance Matrix Adaptation MAP-Annealing