9 papers
Two Routes to Interoceptive Prioritisation: Dissociating Likelihood Precision and Preference Reweighting in Active Inference
St John Grimbly, Nicolas Kuske, Evert A. Boonstra +6
An embodied agent must prioritise among competing bodily needs. In active inference, it can respond to an urgent need by increasing the precision of information about that need or…
Inferring Affective Consciousness in an Artificial Agent: A Case Study
Mark Solms, St John Grimbly, Bruce Bassett +7
Creatures that display 'hedonic place preference behaviour' are thought by many scientists to experience feelings, on the assumption that their attraction to pleasure-producing sub…
Interoceptive Attention as Dynamic Homeostatic Prioritization in a Foraging Agent
St John Grimbly, Nicolas Kuske, Evert A. Boonstra +7
Biological systems must regulate competing needs under limited perceptual bandwidth, where sharpening one estimate costs the capacity to sharpen the others. Any fixed-budget system…
Intention Inference Under Execution Noise: Separating Aleatoric and Epistemic Uncertainty in Social Dilemmas
Kival Mahadew, Jonathan Shock
In noisy social dilemmas, intended actions are stochastically corrupted before execution, so an observed defection may reflect hostile intent or action error. Standard Markov Decis…
Spinning Billiards and Chaos
Jacob S. Lund, Jeff Murugan, Jonathan P. Shock
We investigate the impact of internal spin on chaos in billiard systems. Extending the standard point-particle billiard by coupling translational and rotational degrees of freedom…
Opportunities of Reinforcement Learning in South Africa's Just Transition
Claude Formanek, Callum Rhys Tilbury, Jonathan P. Shock
South Africa stands at a crucial juncture, grappling with interwoven socio-economic challenges such as poverty, inequality, unemployment, and the looming climate crisis. The govern…
Putting Data at the Centre of Offline Multi-Agent Reinforcement Learning
Claude Formanek, Louise Beyers, Callum Rhys Tilbury +2
Offline multi-agent reinforcement learning (MARL) is an exciting direction of research that uses static datasets to find optimal control policies for multi-agent systems. Though th…
Coordination Failure in Cooperative Offline MARL
Callum Rhys Tilbury, Claude Formanek, Louise Beyers +2
Offline multi-agent reinforcement learning (MARL) leverages static datasets of experience to learn optimal multi-agent control. However, learning from static data presents several…
Selective Reincarnation: Offline-to-Online Multi-Agent Reinforcement Learning
Claude Formanek, Callum Rhys Tilbury, Jonathan Shock +2
'Reincarnation' in reinforcement learning has been proposed as a formalisation of reusing prior computation from past experiments when training an agent in an environment. In this…