PlatoLTL: Learning to Generalize Across Symbols in LTL Instructions for Multi-Task RL
Fuente:
arXiv
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866910197274902528 |
|---|---|
| author | Cloete, Jacques Jackermeier, Mathias Havoutis, Ioannis Abate, Alessandro |
| author_facet | Cloete, Jacques Jackermeier, Mathias Havoutis, Ioannis Abate, Alessandro |
| contents | A central challenge in multi-task reinforcement learning (RL) is to train generalist policies capable of performing tasks not seen during training. To facilitate such generalization, linear temporal logic (LTL) has emerged as a powerful formalism for specifying structured, temporally extended tasks to RL agents. While existing approaches to LTL-guided multi-task RL demonstrate generalization across LTL specifications, they are unable to generalize to unseen vocabularies of propositions (or "symbols"), which describe high-level events in LTL. We present PlatoLTL, a novel approach that enables policies to zero-shot generalize not only compositionally across LTL structures, but also parametrically across propositions. We model propositions as parameterized instances of atomic predicates, allowing policies to learn shared structure across related propositions. We propose a novel architecture that embeds and composes parameterized propositions to represent LTL formulae, and demonstrate zero-shot generalization in a range of challenging environments. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_22891 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | PlatoLTL: Learning to Generalize Across Symbols in LTL Instructions for Multi-Task RL Cloete, Jacques Jackermeier, Mathias Havoutis, Ioannis Abate, Alessandro Machine Learning A central challenge in multi-task reinforcement learning (RL) is to train generalist policies capable of performing tasks not seen during training. To facilitate such generalization, linear temporal logic (LTL) has emerged as a powerful formalism for specifying structured, temporally extended tasks to RL agents. While existing approaches to LTL-guided multi-task RL demonstrate generalization across LTL specifications, they are unable to generalize to unseen vocabularies of propositions (or "symbols"), which describe high-level events in LTL. We present PlatoLTL, a novel approach that enables policies to zero-shot generalize not only compositionally across LTL structures, but also parametrically across propositions. We model propositions as parameterized instances of atomic predicates, allowing policies to learn shared structure across related propositions. We propose a novel architecture that embeds and composes parameterized propositions to represent LTL formulae, and demonstrate zero-shot generalization in a range of challenging environments. |
| title | PlatoLTL: Learning to Generalize Across Symbols in LTL Instructions for Multi-Task RL |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2601.22891 |