mlsanityData to 7 Oct 2026

Safe Neurosymbolic Learning with Differentiable Symbolic Execution

arXiv 2203.07671 ↗First posted 15 Mar 2022cs.LG cs.AI cs.PLcited by 9

Not on the map. Why?

Abstract (arXiv)

We study the problem of learning worst-case-safe parameters for programs that use neural networks as well as symbolic, human-written code. Such neurosymbolic programs arise in many safety-critical domains. However, because they can use nondifferentiable operations, it is hard to learn their parameters using existing gradient-based approaches to safe learning. Our approach to this problem, Differentiable Symbolic Execution (DSE), samples control flow paths in a program, symbolically constructs worst-case "safety losses" along these paths, and backpropagates the gradients of these losses through program operations using a generalization of the REINFORCE estimator. We evaluate the method on a mix of synthetic tasks and real-world benchmarks. Our experiments show that DSE significantly outperforms the state-of-the-art DiffAI method on these tasks.

Key facts. Cited by 9 papers in the map (data through 7 Oct 2026). Builds on 20 of its 42 references that are in the map. First posted on arXiv 15 Mar 2022.

Builds on

20 of 42 references.

References · 20

  1. 2014Adam: A Method for Stochastic Optimizationcited by 36,960
  2. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,160
  3. 2017Towards Deep Learning Models Resistant to Adversarial Attackscited by 6,326
  4. 2019Certified Adversarial Robustness via Randomized Smoothingcited by 1,324
  5. 2017Reluplex: An Efficient SMT Solver for Verifying Deep Neural Networkscited by 739
  6. 2018A Reductions Approach to Fair Classificationcited by 585
  7. 2019Batch Policy Learning under Constraintscited by 242
  8. 2019Towards Stable and Efficient Training of Verifiably Robust Neural Networkscited by 204
  9. 2018Formal Security Analysis of Neural Networks using Symbolic Intervalscited by 185
  10. 2018Verifiable Reinforcement Learning via Policy Extractioncited by 176
  11. 2020NNV: The Neural Network Verification Tool for Deep Neural Networks and Learning-Enabled Cyber-Physical Systemscited by 128
  12. 2015Gradient Estimation Using Stochastic Computation Graphscited by 120
8 more
  1. 2018Verisig: verifying safety properties of hybrid systems with neural network controllerscited by 83
  2. 2018Neural Lander: Stable Drone Landing Control using Learned Dynamicscited by 82
  3. 2019An Abstraction-Based Framework for Neural Network Verificationcited by 68
  4. 2021Introduction to Neural Network Verificationcited by 63
  5. 2018Formal Verification of Neural Network Controlled Autonomous Systemscited by 41
  6. 2019Optimization and Abstraction: A Synergistic Approach for Analyzing Neural Network Robustnesscited by 40
  7. 2016A Survey of Symbolic Execution Techniquescited by 36
  8. 2018Symbolic Execution for Deep Neural Networkscited by 8

Built on by

Cited by year

Citing papers per year: 2023: 2; 2024: 2; 2025: 5.

9 papers.

  1. 2025Neuro-Symbolic AI in 2024: A Systematic Reviewcited by 36
  2. 2023Smoothing Methods for Automatic Differentiation Across Conditional Branchescited by 6
  3. 2025On the Promise for Assurance of Differentiable Neurosymbolic Reasoning Paradigmscited by 3
  4. 2023Certifiably Robust Reinforcement Learning through Model-Based Abstract Interpretationcited by 2
  5. 2024A Synergistic Approach In Network Intrusion Detection By Neurosymbolic AIcited by 2
  6. 2024Canopy: Property-Driven Learning for Congestion Controlcited by 2
  7. 2025SAIL: Sound Abstract Interpreters with LLMscited by 2
  8. 2025Composing Linear Layers from Irreduciblescited by 1
  9. 2025Scaling Up without Fading Out: Goal-Aware Sparse GNN for RL-based Generalized Planningcited by 0