Variational Meta Reinforcement Learning
ArchivedA research project on meta reinforcement learning, focusing on variational methods for efficient adaptation to new tasks.
Python PyTorch Meta-learning Reinforcement Learning
Software
A collection of sofware projects that I developed during my PhD and afterwards. These projects are related to reinforcement learning, meta-learning, and preference-based learning.
Projects
Social Gym
Activesocial_gym is a Python library that provides a benchmark of Gymnasium environments for reinforcement learning, active learning, and preference-based learning in the context of social robotics. It simulates robot interactions in human-populated environments, gaze control, and 2D track racing scenarios.
Python Gymnasium Reinforcement Learning
Preference-Based Learning: PBARL
ActiveA modular, research-oriented Python library that integrates Reinforcement Learning (specifically Soft Actor-Critic) and Meta-RL (PEARL) with active Preference-Based Learning.
Python PyTorch RLHF Reinforcement Learning