mlsanityData to 14 Oct 2026

Unsupervised Learning of Neurosymbolic Encoders

arXiv 2107.13132 ↗First posted 28 Jul 2021cs.LG cs.AIcited by 8

Not on the map. Why?

Abstract (arXiv)

We present a framework for the unsupervised learning of neurosymbolic encoders, which are encoders obtained by composing neural networks with symbolic programs from a domain-specific language. Our framework naturally incorporates symbolic expert knowledge into the learning process, which leads to more interpretable and factorized latent representations compared to fully neural encoders. We integrate modern program synthesis techniques with the variational autoencoding (VAE) framework, in order to learn a neurosymbolic encoder in conjunction with a standard decoder. The programmatic descriptions from our encoders can benefit many analysis workflows, such as in behavior modeling where interpreting agent actions and movements is important. We evaluate our method on learning latent representations for real-world trajectory data from animal biology and sports analytics. We show that our approach offers significantly better separation of meaningful categories than standard VAEs and leads to practical gains on downstream analysis tasks, such as for behavior classification.

Key facts. Cited by 8 papers in the map (data through 14 Oct 2026). Builds on 45 of its 79 references that are in the map. First posted on arXiv 28 Jul 2021.

Builds on

45 of 79 references.

References · 45

  1. 2014Adam: A Method for Stochastic Optimizationcited by 37,240
  2. 2017Neural Discrete Representation Learningcited by 4,712
  3. 2013Auto-Encoding Variational Bayescited by 4,053
  4. 2013Distributed Representations of Words and Phrases and their Compositionalitycited by 3,474
  5. 2018UMAP: Uniform Manifold Approximation and Projection for Dimension Reductioncited by 2,898
  6. 2016Categorical Reparameterization with Gumbel-Softmaxcited by 2,739
  7. 2016InfoGAN: Interpretable Representation Learning by Information Maximizing Generative Adversarial Netscited by 860
  8. 2020NVAE: A Deep Hierarchical Variational Autoencodercited by 637
  9. 2018Disentangling by Factorisingcited by 616
  10. 2015Generating Sentences from a Continuous Spacecited by 612
  11. 2020VectorNet: Encoding HD Maps and Agent Dynamics from Vectorized Representationcited by 530
  12. 2014Semi-Supervised Learning with Deep Generative Modelscited by 496
33 more
  1. 2020Learning Lane Graph Representations for Motion Forecastingcited by 382
  2. 2018Understanding disentangling in β-VAEcited by 335
  3. 2020A Survey on Neural Network Interpretabilitycited by 291
  4. 2017Toward Controlled Generation of Textcited by 272
  5. 2018Programmatically Interpretable Reinforcement Learningcited by 193
  6. 2018Verifiable Reinforcement Learning via Policy Extractioncited by 177
  7. 2016Variational Lossy Autoencodercited by 176
  8. 2018Isolating Sources of Disentanglement in Variational Autoencoderscited by 167
  9. 2020DreamCoder: Growing generalizable, interpretable knowledge with wake-sleep Bayesian program learningcited by 161
  10. 2016Discrete Variational Autoencoderscited by 125
  11. 2018Disentangled Sequential Autoencodercited by 120
  12. 2014Neural Variational Inference and Learning in Belief Networkscited by 117
  13. 2016Composing graphical models with neural networks for structured representations and fast inferencecited by 113
  14. 2018Fast Decoding in Sequence Models using Discrete Latent Variablescited by 104
  15. 2018Learning Disentangled Joint Continuous and Discrete Representationscited by 83
  16. 2017Learning to Infer Graphics Programs from Hand-Drawn Imagescited by 74
  17. 2019Learning to Infer and Execute 3D Shape Programscited by 68
  18. 2020Guided Variational Autoencoder for Disentanglement Learningcited by 57
  19. 2018Self-Consistent Trajectory Autoencoder: Hierarchical Reinforcement Learning with Trajectory Embeddingscited by 52
  20. 2018HOUDINI: Lifelong Learning as Program Synthesiscited by 38
  21. 2020Learning Differentiable Programs with Admissible Neural Heuristicscited by 32
  22. 2021The Multi-Agent Behavior Dataset: Mouse Dyadic Social Interactionscited by 31
  23. 2017Learning Hierarchical Features from Generative Modelscited by 27
  24. 2020Task Programming: Learning Data Efficient Behavior Representationscited by 27
  25. 2013Mapping the stereotyped behaviour of freely-moving fruit fliescited by 21
  26. 2016Learning recurrent representations for hierarchical behavior modelingcited by 16
  27. 2020Learning Task-General Representations with Generative Neuro-Symbolic Modelingcited by 14
  28. 2016TerpreT: A Probabilistic Programming Language for Program Inductioncited by 13
  29. 2017Program Synthesis using Abstraction Refinementcited by 13
  30. 2019Learning Calibratable Policies using Programmatic Style-Consistencycited by 11
  31. 2020Polarized-VAE: Proximity Based Disentangled Representation Learning for Text Generationcited by 7
  32. 2017Adversarial Information Factorizationcited by 6
  33. 2016Adaptive Neural Compilationcited by 3

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2022: 4; 2023: 1; 2024: 0; 2025: 1; 2026: 2.

8 papers.

  1. 2025Neuro-Symbolic AI in 2024: A Systematic Reviewcited by 38
  2. 2023Neurosymbolic Grounding for Compositional World Modelscited by 14
  3. 2022MABe22: A Multi-Species Multi-Task Benchmark for Learned Representations of Behaviorcited by 12
  4. 2022Neuro-Symbolic Learning: Principles and Applications in Ophthalmologycited by 7
  5. 2022Neurosymbolic Programming for Sciencecited by 7
  6. 2022Open-Source Tools for Behavioral Video Analysis: Setup, Methods, and Developmentcited by 4
  7. 2026Differentiable Rule Induction from Raw Sequence Inputscited by 2
  8. 2026Grounding vs. Compositionality: On the Non-Complementarity of Reasoning in Neuro-Symbolic Systemscited by 0