相关论文: Learning Program Synthesis for Integer Sequences f…

We propose CoT-Self-Instruct, a synthetic data generation method that instructs LLMs to first reason and plan via Chain-of-Thought (CoT) based on given seed tasks, and then generate a new synthetic example of similar quality and complexity.…

This paper presents a novel method for the automated synthesis of probabilistic programs. The starting point is a program sketch representing a finite family of finite-state Markov chains with related but distinct topologies, and a PCTL…

We introduce a set of eight universal Rules of Inference by which computer programs with known properties (axioms) are transformed into new programs with known properties (theorems). Axioms are presented to formalize a segment of Number…

Modern semantic parsers suffer from two principal limitations. First, training requires expensive collection of utterance-program pairs. Second, semantic parsers fail to generalize at test time to new compositions/structures that have not…

We propose a machine learning framework to synthesize reactive controllers for systems whose interactions with their adversarial environment are modeled by infinite-duration, two-player games over (potentially) infinite graphs. Our…

Writing software exploits is an important practice for offensive security analysts to investigate and prevent attacks. In particular, shellcodes are especially time-consuming and a technical challenge, as they are written in assembly…

One of the first steps in learning how to program is reading and tracing existing code. In order to avoid the error-prone task of generating variations of a tracing exercise, our tool Tatsu generates instances of a given code skeleton…

In many scenarios we need to find the most likely program under a local context, where the local context can be an incomplete program, a partial specification, natural language description, etc. We call such problem program estimation. In…

Synthesizing inductive loop invariants is fundamental to automating program verification. In this work, we observe that Large Language Models (such as gpt-3.5 or gpt-4) are capable of synthesizing loop invariants for a class of programs in…

Our research concerns generating imperative programs from Answer Set Programming Specifications. ASP is highly declarative and is ideal for writing specifications. Further with negation-as-failure it is easy to succinctly represent…

We introduce Modelizer - a novel framework that, given a black-box program, learns a model from its input/output behavior using neural machine translation algorithms. The resulting model mocks the original program: Given an input, the model…

We describe an approach to learn, in a term-rewriting setting, function definitions from input/output equations. By confining ourselves to structurally recursive definitions we obtain a fairly fast learning algorithm that often yields…

Programming or scripting languages used in real-world systems are seldom designed with a formal semantics in mind from the outset. Therefore, developing well-founded analysis tools for these systems requires reverse-engineering a formal…

In models to generate program source code from natural language, representing this code in a tree structure has been a common approach. However, existing methods often fail to generate complex code correctly due to a lack of ability to…

Program synthesis from incomplete specifications (e.g. input-output examples) has gained popularity and found real-world applications, primarily due to its ease-of-use. Since this technology is often used in an interactive setting,…

We show how to systematically implement an algorithm in any imperative or functional programming language. The method is based on the premise that it is easy to write down how an algorithm proceeds on a concrete input. This…

The size and complexity of software applications is increasing at an accelerating pace. Source code repositories (along with their dependencies) require vast amounts of labor to keep them tested, maintained, and up to date. As the…

We present a novel benchmark designed to rigorously evaluate the capabilities of large language models (LLMs) in mathematical reasoning and algorithmic code synthesis tasks. The benchmark comprises integer sequence generation tasks sourced…

Embedding words in high-dimensional vector spaces has proven valuable in many natural language applications. In this work, we investigate whether similarly-trained embeddings of integers can capture concepts that are useful for mathematical…

Language Models (LLMs), such as transformer-based neural networks trained on billions of parameters, have become increasingly prevalent in software engineering (SE). These models, trained on extensive datasets that include code…