Recursive types and bounded quantification are prominent features in many modern programming languages, such as Java, C#, Scala, or TypeScript. Unfortunately, the interaction between recursive types, bounded quantification, and subtyping has shown to be problematic in the past. Consequently, defining a simple foundational calculus that combines those features and has desirable properties, such as decidability, transitivity of subtyping, conservativity, and a sound and complete algorithmic formulation, has been a long-time challenge.

This paper shows how to extend $F_{\le}$ with iso-recursive types in a new calculus called $F_{\le}^{\mu}$. $F_{\le}$ is a well-known polymorphic calculus with bounded quantification. In $F_{\le}^{\mu}$, we add iso-recursive types and correspondingly extend the subtyping relation with iso-recursive subtyping using the recently proposed nominal unfolding rules. In addition, we use so-called structural folding/unfolding rules for typing iso-recursive expressions, inspired by the structural unfolding rule proposed by Abadi et al. (1996). The structural rules add expressive power to the more conventional folding/unfolding rules in the literature, and they enable additional applications. We present several results, including: type soundness; transitivity; the conservativity of $F_{\le}^{\mu}$ over $F_{\le}$; and a sound and complete algorithmic formulation of $F_{\le}^{\mu}$. We study two variants of $F_{\le}^{\mu}$. The first one uses an extension of the $\textrm{kernel}~F_{\le}$ (a well-known decidable variant of $F_{\le}$). This extension accepts equivalent rather than equal bounds and is shown to preserve decidable subtyping. The second variant employs the $\textrm{full}~F_{\le}$ rule for bounded quantification and has undecidable subtyping. Moreover, we also study an extension of the kernel version of $F_{\le}^{\mu}$, called $F_{\le\ge}^{\mu\wedge}$, with a form of intersection types and lower bounded quantification. All the properties from the kernel version of $F_{\le}^{\mu}$ are preserved in $F_{\le\ge}^{\mu\wedge}$. All the results in this paper have been formalized in the Coq theorem prover.

Where dual-numbers forward-mode automatic differentiation (AD) pairs each scalar value with its tangent value, dual-numbers reverse-mode AD attempts to achieve reverse AD using a similarly simple idea: by pairing each scalar value with a backpropagator function. Its correctness and efficiency on higher-order input languages have been analysed by Brunel, Mazza and Pagani, but this analysis used a custom operational semantics for which it is unclear whether it can be implemented efficiently. We take inspiration from their use of linear factoring to optimise dual-numbers reverse-mode AD to an algorithm that has the correct complexity and enjoys an efficient implementation in a standard functional language with support for mutable arrays, such as Haskell. Aside from the linear factoring ingredient, our optimisation steps consist of well-known ideas from the functional programming community. We demonstrate the use of our technique by providing a practical implementation that differentiates most of Haskell98. Where previous work on dual numbers reverse AD has required sequentialisation to construct the reverse pass, we demonstrate that we can apply our technique to task-parallel source programs and generate a task-parallel derivative computation.

We present a novel approach to synthesizing recursive functional programs from input–output examples. Synthesizing a recursive function is challenging because recursive subexpressions should be constructed while the target function has not been fully defined yet. We address this challenge by using a new technique we call block-based pruning. A block refers to a recursion- and conditional-free expression (i.e., straight-line code) that yields an output from a particular input. We first synthesize as many blocks as possible for each input–output example, and then we explore the space of recursive programs, pruning candidates that are inconsistent with the blocks. Our method is based on an efficient version space learning, thereby effectively dealing with a possibly enormous number of blocks. In addition, we present a method that uses sampled input–output behaviors of library functions to enable a goal-directed search for a recursive program using the library. We have implemented our approach in a system called Trio and evaluated it on synthesis tasks from prior work and on new tasks. Our experiments show that Trio significantly outperforms prior work.