Deep Hybrid Models: Infer and Plan in a Dynamic World

To determine an optimal plan for complex tasks, one often deals with dynamic and hierarchical relationships between several entities. Traditionally, such problems are tackled with optimal control, which relies on the optimization of cost functions; instead, a recent biologically motivated proposal c...

Full description

Saved in:
Bibliographic Details
Main Authors: Matteo Priorelli, Ivilin Peev Stoianov
Format: Article
Language:English
Published: MDPI AG 2025-05-01
Series:Entropy
Subjects:
Online Access:https://www.mdpi.com/1099-4300/27/6/570
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849432622456897536
author Matteo Priorelli
Ivilin Peev Stoianov
author_facet Matteo Priorelli
Ivilin Peev Stoianov
author_sort Matteo Priorelli
collection DOAJ
description To determine an optimal plan for complex tasks, one often deals with dynamic and hierarchical relationships between several entities. Traditionally, such problems are tackled with optimal control, which relies on the optimization of cost functions; instead, a recent biologically motivated proposal casts planning and control as an inference process. <i>Active inference</i> assumes that action and perception are two complementary aspects of life whereby the role of the former is to fulfill the predictions inferred by the latter. Here, we present an active inference approach that exploits discrete and continuous processing, based on three features: the representation of <i>potential body configurations</i> in relation to the objects of interest; the use of hierarchical relationships that enable the agent to easily interpret and flexibly expand its body schema for tool use; the definition of <i>potential trajectories</i> related to the agent’s intentions, used to infer and plan with dynamic elements at different temporal scales. We evaluate this <i>deep hybrid model</i> on a habitual task: reaching a moving object after having picked a moving tool. We show that the model can tackle the presented task under different conditions. This study extends past work on planning as inference and advances an alternative direction to optimal control.
format Article
id doaj-art-327d58cc62eb48a994fc970e05804437
institution Kabale University
issn 1099-4300
language English
publishDate 2025-05-01
publisher MDPI AG
record_format Article
series Entropy
spelling doaj-art-327d58cc62eb48a994fc970e058044372025-08-20T03:27:18ZengMDPI AGEntropy1099-43002025-05-0127657010.3390/e27060570Deep Hybrid Models: Infer and Plan in a Dynamic WorldMatteo Priorelli0Ivilin Peev Stoianov1Institute of Cognitive Sciences and Technologies, National Research Council of Italy, 35137 Padova, ItalyInstitute of Cognitive Sciences and Technologies, National Research Council of Italy, 35137 Padova, ItalyTo determine an optimal plan for complex tasks, one often deals with dynamic and hierarchical relationships between several entities. Traditionally, such problems are tackled with optimal control, which relies on the optimization of cost functions; instead, a recent biologically motivated proposal casts planning and control as an inference process. <i>Active inference</i> assumes that action and perception are two complementary aspects of life whereby the role of the former is to fulfill the predictions inferred by the latter. Here, we present an active inference approach that exploits discrete and continuous processing, based on three features: the representation of <i>potential body configurations</i> in relation to the objects of interest; the use of hierarchical relationships that enable the agent to easily interpret and flexibly expand its body schema for tool use; the definition of <i>potential trajectories</i> related to the agent’s intentions, used to infer and plan with dynamic elements at different temporal scales. We evaluate this <i>deep hybrid model</i> on a habitual task: reaching a moving object after having picked a moving tool. We show that the model can tackle the presented task under different conditions. This study extends past work on planning as inference and advances an alternative direction to optimal control.https://www.mdpi.com/1099-4300/27/6/570active inferencemotor controldeep hybrid models
spellingShingle Matteo Priorelli
Ivilin Peev Stoianov
Deep Hybrid Models: Infer and Plan in a Dynamic World
Entropy
active inference
motor control
deep hybrid models
title Deep Hybrid Models: Infer and Plan in a Dynamic World
title_full Deep Hybrid Models: Infer and Plan in a Dynamic World
title_fullStr Deep Hybrid Models: Infer and Plan in a Dynamic World
title_full_unstemmed Deep Hybrid Models: Infer and Plan in a Dynamic World
title_short Deep Hybrid Models: Infer and Plan in a Dynamic World
title_sort deep hybrid models infer and plan in a dynamic world
topic active inference
motor control
deep hybrid models
url https://www.mdpi.com/1099-4300/27/6/570
work_keys_str_mv AT matteopriorelli deephybridmodelsinferandplaninadynamicworld
AT ivilinpeevstoianov deephybridmodelsinferandplaninadynamicworld