Corpus ID: 220250447

Deep Involutive Generative Models for Neural MCMC

@article{Spanbauer2020DeepIG,
  title={Deep Involutive Generative Models for Neural MCMC},
  author={Span Spanbauer and C. Freer and Vikash K. Mansinghka},
  journal={ArXiv},
  year={2020},
  volume={abs/2006.15167}
}
  • Span Spanbauer, C. Freer, Vikash K. Mansinghka
  • Published 2020
  • Computer Science, Mathematics
  • ArXiv
  • We introduce deep involutive generative models, a new architecture for deep generative modeling, and use them to define Involutive Neural MCMC, a new approach to fast neural MCMC. An involutive generative model represents a probability kernel G(φ 7→ φ′) as an involutive (i.e., self-inverting) deterministic function f(φ, π) on an enlarged state space containing auxiliary variables π. We show how to make these models volume preserving, and how to use deep volume-preserving involutive generative… CONTINUE READING

    References

    Publications referenced by this paper.
    SHOWING 1-10 OF 24 REFERENCES
    A-NICE-MC: Adversarial Training for MCMC
    47
    NICE: Non-linear Independent Components Estimation
    601
    i-RevNet: Deep Invertible Networks
    121
    Auto-Encoding Variational Bayes
    8798
    Generalizing Hamiltonian Monte Carlo with Neural Networks
    43
    Meta-Learning MCMC Proposals
    10
    Analyzing Inverse Problems with Invertible Neural Networks
    78
    Neural Variational Inference and Learning in Belief Networks
    516
    Density estimation using Real NVP
    885