Skip to yearly menu bar Skip to main content

Workshop: Differentiable Almost Everything: Differentiable Relaxations, Algorithms, Operators, and Simulators

Differentiating Metropolis-Hastings to Optimize Intractable Densities

Gaurav Arya · Ruben Seyer · Frank Schäfer · Kartik Chandra · Alexander Lew · Mathieu Huot · Vikash Mansinghka · Jonathan Ragan-Kelley · Christopher Rackauckas · Moritz Schauer


We develop an algorithm for automatic differentiation of Metropolis-Hastings samplers, allowing us to differentiate through probabilistic inference, even if the model has discrete components within it. Our approach fuses recent advances in stochastic automatic differentiation with traditional Markov chain coupling schemes, providing an unbiased and low-variance gradient estimator. This allows us to apply gradient-based optimization to objectives expressed as expectations over intractable target densities. We demonstrate our approach by finding an ambiguous observation in a Gaussian mixture model and by maximizing the specific heat in an Ising model.

Chat is not available.