Category Archives: Uncategorized

Probabilistic Programming for Robotics

Autonomous agents use noisy sensors and actuators to interact with a complex external world. As such, inference engines and point estimators are essential to making modern robots work. Often, these are highly specialized and optimized to run in real-time on … Continue reading

Posted in Uncategorized | Comments Off on Probabilistic Programming for Robotics

Contextual Equivalence for a Probabilistic Language with Continuous Random Variables and Recursion

We present a complete reasoning principle for contextual equivalence in an untyped probabilistic programming language. The language includes continuous random variables, conditionals, and scoring. The language also includes recursion, since in an untyped language the standard call-by-value fixpoint combinator is expressible. The language is similar … Continue reading

Posted in Uncategorized | Comments Off on Contextual Equivalence for a Probabilistic Language with Continuous Random Variables and Recursion

S-finite Kernels and Game Semantics for Probabilistic Programming

Staton has recently argued convincingly that first-order functional probabilistic programs with sampling from continuous distributions and soft constraints correspond precisely to so-called s-finite kernels. This class of possibly infinite kernels is a little studied extension of their better known σ-finite cousins, … Continue reading

Posted in Uncategorized | Comments Off on S-finite Kernels and Game Semantics for Probabilistic Programming

Constructive probabilistic semantics with non-spatial locales

Ideally, a probabilistic programming language should admit a computable semantics. But languages often provide operators that denote uncomputable functions, such as comparison of real numbers. While the use of these uncomputable operators may result in uncomputable programs, a programmer can productively … Continue reading

Posted in Uncategorized | Comments Off on Constructive probabilistic semantics with non-spatial locales

TensorFlow Distributions

The TensorFlow Distributions library implements a vision of probability theory adapted to the modern deep-learning paradigm of end-to-end differentiable computation. Building on two basic abstractions, it offers flexible building blocks for probabilistic computation. Distributions provide fast, numerically stable methods for … Continue reading

Posted in Uncategorized | Comments Off on TensorFlow Distributions