Pinned Repositories
JPAL-HA
Justified Human Preferences for Active Learning with Hypothetical Actions (JPAL-HA) is an human-in-the-loop algorithm for safe agent learning in safety-critical environments. It builds on the Parenting algorithm, augmenting it with two novel and generalisable ideas: Justifications and Hypothetical Actions.
swc-rf4
ilkaza's Repositories
ilkaza/JPAL-HA
Justified Human Preferences for Active Learning with Hypothetical Actions (JPAL-HA) is an human-in-the-loop algorithm for safe agent learning in safety-critical environments. It builds on the Parenting algorithm, augmenting it with two novel and generalisable ideas: Justifications and Hypothetical Actions.
ilkaza/swc-rf4