Chronocooked: A Benchmark for Implicit Interval Timing in Reinforcement Learning Agents
2026-08-17 • Artificial Intelligence
Artificial Intelligence
AI summaryⓘ
The authors created Chronocooked, a set of cooking tasks designed to test how reinforcement learning agents understand and use timing without being directly told the time. These tasks require the agents to make decisions based on when actions should happen, even though the timing information is hidden. They kept the environment simple to allow easy testing and to align with how biological systems might work. The authors also tested different types of learning models and highlighted the challenges these agents face with timing, suggesting that teaching agents to perceive time is important for better interactions with humans.
reinforcement learninginterval timingtemporal decision makingreward functionbiologically plausible modelnon-recurrent modelrecurrent modelhuman-robot interactiontime perceptionbenchmark suite
Authors
Amrapali Pednekar, Alvaro Garrido-Perez, Yara Khaluf, Pieter Simoens
Abstract
This paper presents Chronocooked, a reinforcement learning (RL) benchmark suite for studying implicit interval timing in RL agents. Inspired by Overcooked, the suite comprises cooking scenarios that require temporal decision making. The tasks and reward functions are designed such that temporal information is unobserved yet critical for optimal performance. The environment is intentionally kept simple to enable controlled experiments and support biologically plausible models. Evaluation metrics are designed to expose limitations in timing abilities of RL agents, and we report baselines using a non-recurrent, a recurrent, and a biologically plausible model. This work ultimately aims to underscore the need to incorporate time perception and temporal processing in artificial agents designed for human robot interaction and deployment in time dependent human societies.