Searcharxiv⌕ Search
Home/Archive/arXiv

arXiv · 2507.13290

Towards Formal Verification of LLM-Generated Code from Natural Language Prompts

Aaron Councilman·David Jiahao Fu·Aryan Gupta·Chengxiao Wang·David Grove·Yu-Xiong Wang·Vikram Adve
View original source ↗Read source document ↗Export metadata ↓

Abstract

In the past few years LLMs have emerged as a tool that can aid programmers by taking natural language descriptions and generating code based on it. However, the reliability of LLM code generation and current validation techniques for it are far from strong enough to be used for mission-critical or safety-critical applications. In this work we explore ways to offer formal guarantees of correctness to LLM generated code; such guarantees could improve the quality of general AI Code Assistants and support their use for critical applications. To address this challenge we propose to incorporate a Formal Query Language that can represent a user's intent in a formally defined but natural language-like manner that a user can confirm matches their intent. We then have a formal specification of the user intent which we can use to verify that LLM-generated code matches the user's intent. We implement these ideas in our system, Astrogator, for the Ansible programming language, widely used for system administration, including for critical systems. The system includes an intuitive formal query language, a calculus for representing the behavior of Ansible programs, and a symbolic interpreter and a unification algorithm which together are used for the verification. A key innovation in Astrogator is the use of a Knowledge Base to capture system-specific implementation dependencies that greatly reduce the need for system knowledge in expressing formal queries. On a benchmark suite of 21 code-generation tasks, our verifier is able to verify correct code in 83% of cases and identify incorrect code in 92%.

Explore related subjects

cs.PLcs.AI

Keep this discovery

BibTeXRIS

Aaron Councilman, David Jiahao Fu, Aryan Gupta, Chengxiao Wang, David Grove, Yu-Xiong Wang, Vikram Adve. 2025-07-17. Towards Formal Verification of LLM-Generated Code from Natural Language Prompts. https://arxiv.org/abs/2507.13290

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

UnsafeChecker: Finding Soundness Bugs in Rust Safe Abstractions

Rust guarantees memory safety without garbage collection through a strict ownership and borrowing system. However, for low-level systems programming, many widely used libraries rely on the unsafe keyword. These libraries encapsulate raw-pointer operations behind safe APIs to form safe abstractions. A single mistake in this internal unsafe code can break its safety contract, rendering the abstraction unsound and allowing safe clients to trigger undefined behavior. Detecting these potential soundness violations is challenging. Existing static analysis tools for C/C++ ignore Rust-specific safety contracts, while current Rust tools lack the deep semantic modeling required to track the contexts that raw pointers erase. To address this gap, we present UnsafeChecker, a compiler-integrated static analysis framework for detecting potential soundness violations in Rust safe abstractions. UnsafeChecker analyzes Rust MIR using a flow-sensitive abstract interpretation that maintains a shared state with three components: ownership, object validity, and layout. Each warning rule consumes the subset of facts needed for the corresponding Rust safety obligation. UnsafeChecker reports both instruction-level undefined behavior and boundary-level contract violations that may escape through safe APIs. We evaluate UnsafeChecker on a benchmark of 46 RustSec vulnerabilities, which contain 53 ground-truth bugs. UnsafeChecker outperforms several state-of-the-art tools, detecting 32 CVEs and covering 36 bugs (67.9% recall) with 51.6% alert-level precision. Furthermore, in a large-scale scan of real-world crates on crates.io, UnsafeChecker uncovered 114 previously unknown bugs across 83 crates, with 45 confirmed and 27 already fixed by maintainers.

cs.PL↗
arXiv2026

Mapping Dynamic, Hierarchical Quantum Circuits

Qubit mapping is a critical pass in quantum compilation. Despite various advances, dynamic circuits, those exhibiting data dependent control-flow, often resulting from qubit measurements, are not yet supported by the vast majority of available qubit mappers. The crucial limitation to overcome is the dependence on flat, one-dimensional representations of circuits. Further, qubit mappers currently lack compiler abstractions that capture the hierarchical nature of circuits, hindering the qubit mapping process. In this paper, 1 we introduce a new qubit mapping method and analyses to tackle hierarchical dynamic circuits. Our novelty resides in four key aspects: modeling (statically) sub-circuits in disjoint control-flow paths, introducing a novel Qubit Reconciliation pass to maintain consistency between sub-circuit and control-flow boundaries, a loop-entry remapping pass, and a refined cost function enhanced for SWAP count, circuit depth, circuit latency and error. We demonstrate the efficiency of our approach on a wide range of dynamic circuits on two monolithic Quantum Processing Units of 127 and 156 qubits, and on chiplet hexagon-based QPUs. On monolithic QPUs, our qubit mapper improves the SWAP count by up to 52%, depth by up to 18%, latency by up to 18.6%, and error by up to 40%. On chiplet architectures, we achieve improvements of up to 36% on SWAP count, 8.7% on depth, 15% on latency, and 15% of error.

cs.PL↗
arXiv2026

Erased Postulates, Identity Types and Quotients

This text is concerned with the question of whether, in type theory with erasure annotations, one can postulate that some type is inhabited and still have a guarantee that a program will not get stuck. Previous work has provided such guarantees for consistent erased postulates, i.e. postulates that are restricted to be used in erased contexts. Here those guarantees are extended to type theory with identity types. Similar ideas provide a simple way to support quotient types: it is shown that one can let things like "the equivalence classes for two related values are equal" be erased postulates and have an eliminator that only computes for the equivalence class constructor, and still get a guarantee that programs will compute correctly. Another question is whether programs compute correctly if one is allowed to transport (cast) using erased identity proofs. It is shown that this is safe in the absence of quotients and postulates, and in the presence of quotients and erased postulates that can be implemented using equality reflection. However, unrestricted transports of this kind are not compatible with erased, postulated univalence. For that reason the text includes a study of the function []-cong, which encapsulates a limited form of transport for erased identity proofs. The text is accompanied by machine-checked Agda proofs.

cs.PL↗