arXiv subjects

Tiark Rompf

Publications and source records attributed to Tiark Rompf.

22 records · Page 2Linked to original sources

arXiv2018

From Gameplay to Symbolic Reasoning: Learning SAT Solver Heuristics in the Style of Alpha(Go) Zero

Despite the recent successes of deep neural networks in various fields such as image and speech recognition, natural language processing, and reinforcement learning, we still face big challenges in bringing the power of numeric optimization to symbolic reasoning. Researchers have proposed different avenues such as neural machine translation for proof synthesis, vectorization of symbols and expressions for representing symbolic patterns, and coupling of neural back-ends for dimensionality reduction with symbolic front-ends for decision making. However, these initial explorations are still only point solutions, and bear other shortcomings such as lack of correctness guarantees. In this paper, we present our approach of casting symbolic reasoning as games, and directly harnessing the power of deep reinforcement learning in the style of Alpha(Go) Zero on symbolic problems. Using the Boolean Satisfiability (SAT) problem as showcase, we demonstrate the feasibility of our method, and the advantages of modularity, efficiency, and correctness guarantees.

cs.AI↗

arXiv2017

Flare: Native Compilation for Heterogeneous Workloads in Apache Spark

The need for modern data analytics to combine relational, procedural, and map-reduce-style functional processing is widely recognized. State-of-the-art systems like Spark have added SQL front-ends and relational query optimization, which promise an increase in expressiveness and performance. But how good are these extensions at extracting high performance from modern hardware platforms? While Spark has made impressive progress, we show that for relational workloads, there is still a significant gap compared with best-of-breed query engines. And when stepping outside of the relational world, query optimization techniques are ineffective if large parts of a computation have to be treated as user-defined functions (UDFs). We present Flare: a new back-end for Spark that brings performance closer to the best SQL engines, without giving up the added expressiveness of Spark. We demonstrate order of magnitude speedups both for relational workloads such as TPC-H, as well as for a range of machine learning kernels that combine relational and iterative functional processing. Flare achieves these results through (1) compilation to native code, (2) replacing parts of the Spark runtime system, and (3) extending the scope of optimization and code generation to large classes of UDFs.

cs.DB↗

arXiv2016

From F to DOT: Type Soundness Proofs with Definitional Interpreters

Scala's type system unifies ML modules, object-oriented, and functional programming. The Dependent Object Types (DOT) family of calculi has been proposed as a new foundation for Scala and similar languages. Unfortunately, it is not clear how DOT relates to any well-known type systems, and type soundness has only been established for very restricted subsets. In fact, important Scala features are known to break at least one key metatheoretic property such as environment narrowing or subtyping transitivity, which are usually required for a type soundness proof. First, and, perhaps surprisingly, we show how rich DOT calculi can still be proved sound. The key insight is that narrowing and subtyping transitivity only need to hold for runtime objects, but not for code that is never executed. Alas, the dominant method of proving type soundness, Wright and Felleisen's syntactic approach, is based on term rewriting, which does not a priori make a distinction between runtime and type assignment time. Second, we demonstrate how type soundness can be proved for advanced, polymorphic, type systems with respect to high-level, definitional interpreters, implemented in Coq. We present the first mechanized soundness proof in this style for System F<: and several extensions, including mutable references. Our proofs use only simple induction: another surprising result, as the combination of big-step semantics, mutable references, and polymorphism is commonly believed to require co-inductive proof techniques. Third, we show how DOT-like calculi emerge as generalizations of F<:, exposing a rich design space of calculi with path-dependent types which we collectively call System D. Armed with insights from the definitional interpreter semantics, we also show how equivalent small-step semantics and soundness proofs in Wright-Felleisen-style can be derived for these systems.

cs.PL↗

arXiv2011

Building-Blocks for Performance Oriented DSLs

Domain-specific languages raise the level of abstraction in software development. While it is evident that programmers can more easily reason about very high-level programs, the same holds for compilers only if the compiler has an accurate model of the application domain and the underlying target platform. Since mapping high-level, general-purpose languages to modern, heterogeneous hardware is becoming increasingly difficult, DSLs are an attractive way to capitalize on improved hardware performance, precisely by making the compiler reason on a higher level. Implementing efficient DSL compilers is a daunting task however, and support for building performance-oriented DSLs is urgently needed. To this end, we present the Delite Framework, an extensible toolkit that drastically simplifies building embedded DSLs and compiling DSL programs for execution on heterogeneous hardware. We discuss several building blocks in some detail and present experimental results for the OptiML machine-learning DSL implemented on top of Delite.

cs.PL↗