mlsanityData to 7 Oct 2026

pix2rule: End-to-end Neuro-symbolic Rule Learning

arXiv 2106.07487 ↗First posted 14 Jun 2021cs.LG cs.AIcited by 9

Not on the map. Why?

Abstract (arXiv)

Humans have the ability to seamlessly combine low-level visual input with high-level symbolic reasoning often in the form of recognising objects, learning relations between them and applying rules. Neuro-symbolic systems aim to bring a unifying approach to connectionist and logic-based principles for visual processing and abstract reasoning respectively. This paper presents a complete neuro-symbolic method for processing images into objects, learning relations and logical rules in an end-to-end fashion. The main contribution is a differentiable layer in a deep learning architecture from which symbolic relations and rules can be extracted by pruning and thresholding. We evaluate our model using two datasets: subgraph isomorphism task for symbolic rule learning and an image classification domain with compound relations for learning objects, relations and rules. We demonstrate that our model scales beyond state-of-the-art symbolic learners and outperforms deep relational neural network architectures.

Key facts. Cited by 9 papers in the map (data through 7 Oct 2026). Builds on 41 of its 76 references that are in the map. First posted on arXiv 14 Jun 2021.

Builds on

41 of 76 references.

References · 41

  1. 2014Adam: A Method for Stochastic Optimizationcited by 36,960
  2. 2014Neural Machine Translation by Jointly Learning to Align and Translatecited by 4,530
  3. 2013Distributed Representations of Words and Phrases and their Compositionalitycited by 3,462
  4. 2016Categorical Reparameterization with Gumbel-Softmaxcited by 2,708
  5. 2018Relational inductive biases, deep learning, and graph networkscited by 1,341
  6. 2014On the Properties of Neural Machine Translation: Encoder-Decoder Approachescited by 1,303
  7. 2016The Concrete Distribution: A Continuous Relaxation of Discrete Random Variablescited by 1,135
  8. 2020Object-Centric Learning with Slot Attentioncited by 865
  9. 2014Neural Turing Machinescited by 630
  10. 2019Neural Probabilistic Logic Programming in DeepProbLogcited by 389
  11. 2019MONet: Unsupervised Scene Decomposition and Representationcited by 388
  12. 2019The Neuro-Symbolic Concept Learner: Interpreting Scenes, Words, and Sentences From Natural Supervisioncited by 369
29 more
  1. 2014Memory Networkscited by 354
  2. 2017A simple neural network module for relational reasoningcited by 345
  3. 2018Neural-Symbolic VQA: Disentangling Reasoning from Vision and Language Understandingcited by 304
  4. 2017Learning Explanatory Rules from Noisy Datacited by 210
  5. 2019Graph Matching Networks for Learning the Similarity of Graph Structured Objectscited by 202
  6. 2017Neural-Symbolic Learning and Reasoning: A Survey and Interpretationcited by 195
  7. 2017End-to-End Differentiable Provingcited by 159
  8. 2019Neural Logic Machinescited by 144
  9. 2019An Attentive Survey of Attention Modelscited by 123
  10. 2020Logical Neural Networkscited by 118
  11. 2016Neural Symbolic Machines: Learning Semantic Parsers on Freebase with Weak Supervisioncited by 101
  12. 2015Neural Programmer-Interpreterscited by 96
  13. 2015Inferring Algorithmic Patterns with Stack-Augmented Recurrent Netscited by 76
  14. 2021Neural Production Systems: Learning Rule-Governed Visual Dynamicscited by 75
  15. 2017Classical Planning in Deep Latent Space: Bridging the Subsymbolic-Symbolic Boundarycited by 58
  16. 2020Learning Physical Graph Representations from Visual Scenescited by 50
  17. 2018From Word to Sense Embeddings: A Survey on Vector Representations of Meaningcited by 49
  18. 2019An Explicitly Relational Neural Network Architecturecited by 35
  19. 2016TensorLog: A Differentiable Deductive Databasecited by 29
  20. 2020The ILASP system for Inductive Learning of Answer Set Programscited by 27
  21. 2021Differentiable Inductive Logic Programming for Structured Examplescited by 26
  22. 2019Inductive Logic Programming via Differentiable Deep Neural Logic Networkscited by 25
  23. 2019Unsupervised Grounding of Plannable First-Order Logic Representation from Imagescited by 24
  24. 2019Learning Algorithms via Neural Logic Networkscited by 17
  25. 2015Learning Simple Algorithms from Examplescited by 12
  26. 2020Neural Datalog Through Time: Informed Temporal Modeling via Logical Specificationcited by 10
  27. 2018DeepLogic: Towards End-to-End Differentiable Logical Reasoningcited by 8
  28. 2015Lifted Relational Neural Networkscited by 5
  29. 2020NSL: Hybrid Interpretable Learning From Noisy Raw Datacited by 1

Built on by

Cited by year

Citing papers per year: 2023: 2; 2024: 3; 2025: 4.

9 papers.

  1. 2025Neuro-Symbolic AI in 2024: A Systematic Reviewcited by 36
  2. 2024TELLER: A Trustworthy Framework for Explainable, Generalizable and Controllable Fake News Detectioncited by 10
  3. 2024Neuro-Symbolic AI: Explainability, Challenges, and Future Trendscited by 10
  4. 2023Neuro-symbolic Rule Learning in Real-world Classification Taskscited by 7
  5. 2025Neural DNF-MT: A Neuro-symbolic Approach for Learning Interpretable and Editable Policiescited by 4
  6. 2023Learning Type-Generalized Actions for Symbolic Planningcited by 2
  7. 2025Neural Logic Networks for Interpretable Classificationcited by 2
  8. 2024Bridging the Gap: Representation Spaces in Neuro-Symbolic AIcited by 0
  9. 2025Disentangling Neural Disjunctive Normal Form Modelscited by 0