mlsanityData to 14 Oct 2026

Building Intelligent Agents with Neuro-Symbolic Concepts

arXiv 2505.06191 ↗First posted 9 May 2025cs.AI cs.CL cs.CVcited by 10

Not on the map. Why?

Abstract (arXiv)

This article presents a concept-centric paradigm for building agents that can learn continually and reason flexibly. The concept-centric agent utilizes a vocabulary of neuro-symbolic concepts. These concepts, such as object, relation, and action concepts, are grounded on sensory inputs and actuation outputs. They are also compositional, allowing for the creation of novel concepts through their structural combination. To facilitate learning and reasoning, the concepts are typed and represented using a combination of symbolic programs and neural network representations. Leveraging such neuro-symbolic concepts, the agent can efficiently learn and recombine them to solve various tasks across different domains, ranging from 2D images, videos, 3D scenes, and robotic manipulation tasks. This concept-centric framework offers several advantages, including data efficiency, compositional generalization, continual learning, and zero-shot transfer.

Key facts. Cited by 10 papers in the map (data through 14 Oct 2026). Builds on 26 of its 42 references that are in the map. First posted on arXiv 9 May 2025.

Builds on

26 of 42 references.

References · 26

  1. 2016CLEVR: A Diagnostic Dataset for Compositional Language and Elementary Visual Reasoningcited by 1,457
  2. 2023ViperGPT: Visual Inference via Python Execution for Reasoningcited by 603
  3. 2019Neural Probabilistic Logic Programming in DeepProbLogcited by 394
  4. 2015Neural Module Networkscited by 376
  5. 2019The Neuro-Symbolic Concept Learner: Interpreting Scenes, Words, and Sentences From Natural Supervisioncited by 372
  6. 2019SATNet: Bridging deep learning and logical reasoning using a differentiable satisfiability solvercited by 173
  7. 2020DreamCoder: Growing generalizable, interpretable knowledge with wake-sleep Bayesian program learningcited by 161
  8. 2023NeurASP: Embracing Neural Networks into Answer Set Programmingcited by 92
  9. 2021Grounding Physical Concepts of Objects and Events Through Dynamic Visual Reasoningcited by 68
  10. 2023NS3D: Neuro-Symbolic Grounding of 3D Objects and Relationscited by 57
  11. 2023Interpretable Neural-Symbolic Concept Reasoningcited by 57
  12. 2021Dynamic Visual Reasoning by Learning Differentiable Physics Models from Video and Languagecited by 55
14 more
  1. 2023Safe Reinforcement Learning via Probabilistic Logic Shieldscited by 42
  2. 2020Visual Concept-Metaconcept Learningcited by 39
  3. 2020Neuro-Symbolic Visual Reasoning: Disentangling "Visual" from "Reasoning"cited by 35
  4. 2023Motion Question Answering via Modular Motion Programscited by 27
  5. 2023Programmatically Grounded, Compositionally Generalizable Robotic Manipulationcited by 26
  6. 2016Lifted Rule Injection for Relation Embeddingscited by 22
  7. 2021Learning with Algorithmic Supervision via Continuous Relaxationscited by 21
  8. 2022FALCON: Fast Visual Concept Learning by Integrating Images, Linguistic descriptions, and Conceptual Relationscited by 20
  9. 2020A Competence-aware Curriculum for Visual Concepts Learning via Question Answeringcited by 16
  10. 2020Disentangling 3D Prototypical Networks For Few-Shot Concept Learningcited by 15
  11. 2023Scalable Neural-Probabilistic Answer Set Programmingcited by 14
  12. 2023Learning Differentiable Logic Programs for Abstract Visual Reasoningcited by 12
  13. 2024What Makes a Maze Look Like a Maze?cited by 11
  14. 2022Learning Neuro-symbolic Programs for Language Guided Robot Manipulationcited by 10

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 2; 2026: 8.

10 papers.

  1. 2026Lifting Traces to Logic: Programmatic Skill Induction with Neuro-Symbolic Learning for Long-Horizon Agentic Taskscited by 9
  2. 2026Neuro-Symbolic Compliance: Integrating LLMs and SMT Solvers for Automated Financial Legal Analysiscited by 3
  3. 2025DR. WELL: Dynamic Reasoning and Learning with Symbolic World Model for Embodied LLM-Based Multi-Agent Collaborationcited by 2
  4. 2025Beyond Either-Or Reasoning: Transduction and Induction as Cooperative Problem-Solving Paradigmscited by 1
  5. 2026Visual General Intelligence: A White Papercited by 1
  6. 2026REASON: Accelerating Probabilistic Logical Reasoning for Scalable Neuro-Symbolic Intelligencecited by 0
  7. 2026Neuro-Symbolic Decoding of Neural Activitycited by 0
  8. 2026GRAIL: Autonomous Concept Grounding for Neuro-Symbolic Reinforcement Learningcited by 0
  9. 2026Agentic Neuro-Symbolic Planning and Commissioning for Human-in-the-Loop Industrial Robotics with Digital Twinscited by 0
  10. 2026Neuro-Symbolic Agentic AI for Networked Low-Altitude UAVscited by 0