Reinforcement Learning

Reinforcement Learning (RL) research often models environments as Markov decision processes. Yet many real-world tasks are …

We describe a research programme aimed at the construction of a single, generally intelligent agent—one competent across all …

Safe offline RL typically assumes access to dense per-step cost annotations, but in practice supervisors provide only trajectory-level …

Reinforcement learning is often framed around episodic, discounted, or average scalar rewards. While useful, these views miss a core …

An important problem in reinforcement learning is designing agents that learn to solve tasks safely in an environment. A common …

One avenue for creating a generally intelligent agent is to equip it with the ability to combine its previously learned behaviours to …

Evaluating moral alignment in agents navigating conflicting, hierarchically structured human norms is a critical challenge at the …

We consider how to construct state abstractions compatible with a given set of abstract actions, to obtain a well-formed abstract …

Skill composition is a growing area of interest within Reinforcement Learning (RL) research. For example, if designing a robot for …

An important property of long-lived agents is the ability to reuse existing knowledge to solve new tasks. An appealing approach towards …