Policy Rehearsing: Training Generalizable Policies for Reinforcement Learning
Human beings can make adaptive decisions in a preparatory manner, ie, by making
preparations in advance, which offers significant advantages in scenarios where both online …
preparations in advance, which offers significant advantages in scenarios where both online …
No regrets: Investigating and improving regret approximations for curriculum discovery
What data or environments to use for training to improve downstream performance is a
longstanding and very topical question in reinforcement learning. In particular …
longstanding and very topical question in reinforcement learning. In particular …
minimax: Efficient Baselines for Autocurricula in JAX
Unsupervised environment design (UED) is a form of automatic curriculum learning for
training robust decision-making agents to zero-shot transfer into unseen environments. Such …
training robust decision-making agents to zero-shot transfer into unseen environments. Such …
Environment curriculum generation via large language models
Recent work has demonstrated that a promising strategy for teaching robots a wide range of
complex skills is by training them on a curriculum of progressively more challenging …
complex skills is by training them on a curriculum of progressively more challenging …
The Overcooked Generalisation Challenge
We introduce the Overcooked Generalisation Challenge (OGC)-the first benchmark to study
agents' zero-shot cooperation abilities when faced with novel partners and levels in the …
agents' zero-shot cooperation abilities when faced with novel partners and levels in the …
Multi-Agent Diagnostics for Robustness via Illuminated Diversity
In the rapidly advancing field of multi-agent systems, ensuring robustness in unfamiliar and
adversarial settings is crucial. Notwithstanding their outstanding performance in familiar …
adversarial settings is crucial. Notwithstanding their outstanding performance in familiar …
Syllabus: Portable Curricula for Reinforcement Learning Agents
Curriculum learning has been a quiet yet crucial component of many of the high-profile
successes of reinforcement learning. Despite this, none of the major reinforcement learning …
successes of reinforcement learning. Despite this, none of the major reinforcement learning …
Refining Minimax Regret for Unsupervised Environment Design
In unsupervised environment design, reinforcement learning agents are trained on
environment configurations (levels) generated by an adversary that maximises some …
environment configurations (levels) generated by an adversary that maximises some …
Adversarial Environment Design via Regret-Guided Diffusion Models
Training agents that are robust to environmental changes remains a significant challenge in
deep reinforcement learning (RL). Unsupervised environment design (UED) has recently …
deep reinforcement learning (RL). Unsupervised environment design (UED) has recently …
Scenario-Based Curriculum Generation for Multi-Agent Autonomous Driving
The automated generation of diverse and complex training scenarios has been an important
ingredient in many complex learning tasks. Especially in real-world application domains …
ingredient in many complex learning tasks. Especially in real-world application domains …