Grounding LTL Tasks in Sub-Symbolic RL Environments for Zero-Shot Generalization

Matteo Pannacci, Andrea Fanti, Elena Umili, Roberto Capobianco; 7, 2026.

Presented at the Reinforcement Learning Conference (RLC), Montréal, Quebec, Canada, August 15–17, 2026.

Abstract

In this work we address the problem of training a Reinforcement Learning agent to follow multiple temporally-extended instructions expressed in Linear Temporal Logic in sub-symbolic environments. Previous multi-task work has mostly relied on knowledge of the mapping between raw observations and symbols appearing in the formulae. We drop this unrealistic assumption by jointly training a multi-task policy and a symbol grounder with the same experience. The symbol grounder is trained only from raw observations and sparse rewards via Neural Reward Machines in a semi-supervised fashion. Experiments on vision-based environments show that our method achieves performance comparable to using the true symbol grounding and significantly outperforms state-of-the-art methods for sub-symbolic environments.

[abs][pdf]

BibTeX

@article{pannacci2026grounding,
    title={Grounding LTL Tasks in Sub-Symbolic RL Environments for Zero-Shot Generalization},
    author={Matteo Pannacci and Andrea Fanti and Elena Umili and Roberto Capobianco},
    journal={Reinforcement Learning Journal},
    volume={7},
    pages={},
    year={2026}
}