mlsanityData to 7 Oct 2026

REBAR: Low-variance, unbiased gradient estimates for discrete latent variable models

arXiv 1703.07370 ↗First posted 21 Mar 2017cs.LG stat.MLcited by 78

Not on the map. Why?

Abstract (arXiv)

Learning in models with discrete latent variables is challenging due to high variance gradient estimators. Generally, approaches have relied on control variates to reduce the variance of the REINFORCE estimator. Recent work (Jang et al. 2016, Maddison et al. 2016) has taken a different approach, introducing a continuous relaxation of discrete variables to produce low-variance, but biased, gradient estimates. In this work, we combine the two approaches through a novel control variate that produces low-variance, unbiased gradient estimates. Then, we introduce a modification to the continuous relaxation and show that the tightness of the relaxation can be adapted online, removing it as a hyperparameter. We show state-of-the-art variance reduction on several benchmark generative modeling tasks, generally leading to faster convergence to a better final log-likelihood.

Key facts. Cited by 78 papers in the map (data through 7 Oct 2026). First posted on arXiv 21 Mar 2017.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2019: 1; 2020: 0; 2021: 23; 2022: 17; 2023: 14; 2024: 6; 2025: 6; 2026: 11.

78 papers.

Sort:

By area:

  1. 2022Riemannian Diffusion Modelscited by 120
  2. 2021Deep Adaptive Design: Amortizing Sequential Bayesian Experimental Designcited by 103
  3. 2022GFlowNets and variational inferencecited by 103
  4. 2021Implicit MLE: Backpropagating Through Discrete Exponential Family Distributionscited by 97
  5. 2023Soft Merging of Experts with Adaptive Routingcited by 88
  6. 2021A Review of the Gumbel-max Trick and its Extensions for Discrete Stochasticity in Machine Learningcited by 82
  7. 2021Have We Learned to Explain?: How Interpretability Methods Can Learn to Encode Predictions in their Interpretationscited by 59
  8. 2022Ordered Subgraph Aggregation Networkscited by 57
  9. 2021RDA: Robust Domain Adaptation via Fourier Adversarial Attackingcited by 48
  10. 2021Differentiable Subset Pruning of Transformer Headscited by 46
  11. 2021Locally Sparse Neural Networks for Tabular Biomedical Datacited by 34
  12. 2022SIMPLE: A Gradient Estimator for k-Subset Samplingcited by 34
38 more
  1. 2023Bridging Discrete and Backpropagation: Straight-Through and Beyondcited by 31
  2. 2022Automatic Differentiation of Programs with Discrete Randomnesscited by 28
  3. 2021Probabilistic Mixture-of-Experts for Efficient Deep Reinforcement Learningcited by 26
  4. 2021Mind Mappings: Enabling Efficient Algorithm-Accelerator Mapping Space Searchcited by 24
  5. 2023Neural Probabilistic Logic Programming in Discrete-Continuous Domainscited by 24
  6. 2023Efficient Attention via Control Variatescited by 23
  7. 2022Gradient Estimation with Discrete Stein Operatorscited by 22
  8. 2022Sparse Graph Learning from Spatiotemporal Time Seriescited by 22
  9. 2023Sparse Backpropagation for MoE Trainingcited by 16
  10. 2023Differentiable Sampling of Categorical Distributions Using the CatLog-Derivative Trickcited by 13
  11. 2021A unified view of likelihood ratio and reparameterization gradientscited by 12
  12. 2022Adaptive Perturbation-Based Gradient Estimation for Discrete Latent Variable Modelscited by 12
  13. 2021Storchastic: A Framework for General Stochastic Automatic Differentiationcited by 11
  14. 2021Coupled Gradient Estimators for Discrete Latent Variablescited by 9
  15. 2021ARMS: Antithetic-REINFORCE-Multi-Sample Gradient for Binary Variablescited by 8
  16. 2021Double Control Variates for Gradient Estimation in Discrete Latent Variable Modelscited by 8
  17. 2021Leveraging Recursive Gumbel-Max Trick for Approximate Inference in Combinatorial Spacescited by 7
  18. 2022Gradient Estimation for Binary Latent Variables via Gradient Variance Clippingcited by 7
  19. 2022Variational Sparse Coding with Learned Thresholdingcited by 6
  20. 2022Foundation Posteriors for Approximate Probabilistic Inferencecited by 6
  21. 2022Metrizing Fairnesscited by 6
  22. 2021Fitting summary statistics of neural data with a differentiable spiking network simulatorcited by 5
  23. 2022Variational Open-Domain Question Answeringcited by 5
  24. 2023Improving Dual-Encoder Training through Dynamic Indexes for Negative Miningcited by 5
  25. 2021Bias-Variance Tradeoffs in Single-Sample Binary Gradient Estimatorscited by 4
  26. 2021CARMS: Categorical-Antithetic-REINFORCE Multi-Sample Gradient Estimatorcited by 4
  27. 2021Multi-Modal Mutual Information Maximization: A Novel Approach for Unsupervised Deep Cross-Modal Hashingcited by 4
  28. 2021Debiasing a First-order Heuristic for Approximate Bi-level Optimizationcited by 3
  29. 2021Hybrid Memoised Wake-Sleep: Approximate Inference at the Discrete-Continuous Interfacecited by 3
  30. 2024CoRMF: Criticality-Ordered Recurrent Mean Field Ising Solvercited by 3
  31. 2023Revisiting the Gumbel-Softmax in MADDPGcited by 2
  32. 2023Efficient Learning of Discrete-Continuous Computation Graphscited by 2
  33. 2025Bayesian Experimental Design for Model Discrepancy Calibration: An Auto-Differentiable Ensemble Kalman Inversion Approachcited by 2
  34. 2025Mamba-FETrack V2: Revisiting State Space Model for Frame-Event based Visual Object Trackingcited by 2
  35. 2026Exact Discrete Stochastic Simulation with Deep-Learning-Scale Gradient Optimizationcited by 2
  36. 2026On Advantage Estimates for Max@K Policy Gradientscited by 2
  37. 2022VFDS: Variational Foresight Dynamic Selection in Bayesian Neural Networks for Efficient Human Activity Recognitioncited by 1
  38. 2022Latent Variable Modelling Using Variational Autoencoders: A surveycited by 1

50 most cited of 78 All 78 (JSON)