A review of symbolic, subsymbolic and hybrid methods for sequential decision making

C Núñez-Molina, P Mesejo… - ACM Computing …, 2024 - dl.acm.org
In the field of Sequential Decision Making (SDM), two paradigms have historically vied for
supremacy: Automated Planning (AP) and Reinforcement Learning (RL). In the spirit of …

Deepsym: Deep symbol generation and rule learning for planning from unsupervised robot interaction

A Ahmetoglu, MY Seker, J Piater, E Oztop… - Journal of Artificial …, 2022 - jair.org
Symbolic planning and reasoning are powerful tools for robots tackling complex tasks.
However, the need to manually design the symbols restrict their applicability, especially for …

Egocentric planning for scalable embodied task achievement

X Liu, H Palacios, C Muise - Advances in Neural …, 2023 - proceedings.neurips.cc
Embodied agents face significant challenges when tasked with performing actions in diverse
environments, particularly in generalizing across object types and executing suitable actions …

From Reals to Logic and Back: Inventing Symbolic Vocabularies, Actions and Models for Planning from Raw Data

N Shah, J Nagpal, P Verma, S Srivastava - arXiv preprint arXiv …, 2024 - arxiv.org
Hand-crafted, logic-based state and action representations have been widely used to
overcome the intractable computational complexity of long-horizon robot planning problems …

Learning Abstract World Model for Value-preserving Planning with Options

R Rodriguez-Sanchez, G Konidaris - arXiv preprint arXiv:2406.15850, 2024 - arxiv.org
General-purpose agents require fine-grained controls and rich sensory inputs to perform a
wide range of tasks. However, this complexity often leads to intractable decision-making …

[PDF][PDF] A Survey on Model-Free Goal Recognition

L Amado, SP Shainkopf, RF Pereira, R Mirsky… - IJCAI 2024: 33rd …, 2024 - ijcai.org
Goal Recognition is the task of inferring an agent's intentions from a set of observations.
Existing recognition approaches have made considerable advances in domains such as …

Discovering predictive relational object symbols with symbolic attentive layers

A Ahmetoglu, B Celik, E Oztop… - IEEE Robotics and …, 2024 - ieeexplore.ieee.org
In this letter, we propose and realize a new deep learning architecture for discovering
symbolic representations for objects and their relations based on the self-supervised …

Learning multi-object symbols for manipulation with attentive deep effect predictors

A Ahmetoglu, E Oztop, E Ugur - arXiv preprint arXiv:2208.01021, 2022 - arxiv.org
In this paper, we propose a concept learning architecture that enables a robot to build
symbols through self-exploration by interacting with a varying number of objects. Our aim is …

Neuro-Symbolic Learning of Lifted Action Models from Visual Traces

K Xi, S Gould, S Thiébaux - Proceedings of the International Conference …, 2024 - ojs.aaai.org
Abstract Model-based planners rely on action models to describe available actions in terms
of their preconditions and effects. Nonetheless, manually encoding such models is …

AI Planning: A Primer and Survey (Preliminary Report)

DZ Chen, P Verma, S Srivastava, M Katz… - arXiv preprint arXiv …, 2024 - arxiv.org
Automated decision-making is a fundamental topic that spans multiple sub-disciplines in AI:
reinforcement learning (RL), AI planning (AP), foundation models, and operations research …