Zum Inhalt springen

Nur-Lesen-Fassung: Das Forschungsinformationssystem (FIS) ist aus Sicherheitsgründen vorübergehend durch diese Nur-Lesen-Fassung ersetzt. Alle öffentlichen Inhalte und Volltexte sind weiterhin abrufbar; Anmeldung, Bearbeitung, Einreichung und Statistiken stehen nicht zur Verfügung. Bei Fragen: fis@uni-bamberg.de

Start › Publikationen › Neuro-Symbolic Verification of LLM Outputs for Data-Sensitive Domains (extended preprint)

Publikation

Neuro-Symbolic Verification of LLM Outputs for Data-Sensitive Domains (extended preprint)

Sigloch, Paul; Benzmüller, Christoph (2026): Neuro-Symbolic Verification of LLM Outputs for Data-Sensitive Domains (extended preprint), in: arXiv, doi: 10.48550/arxiv.2605.26942.

Details

Fakultät/Lehrstuhl
Autorin(nen)/Autor(en)
Sigloch, Paul
Verlagsangabe
arXiv
Erscheinungsjahr
2026
Seiten
22
Sprache(n)
Englisch
DOI
Abstract

LLMs deployed in high-stakes domains face fundamental reliability challenges: hallucinations, inconsistencies, and privacy vulnerabilities introduce unacceptable risks where errors carry legal, financial, or safety consequences. This paper presents a hybrid verification architecture combining formal symbolic methods with neural semantic analysis to provide complementary guarantees for LLM-generated content. This architecture employs logical reasoning for input verification, leveraging completeness properties to provide decidable guarantees on structured requirements. For output validation, embedding-based semantic similarity detects contextual hallucinations where formal methods lack expressiveness. This separation is realized in a parallel, actor-based pipeline, addressing limitations of prompt-based self-verification approaches, which inherit the distributional biases that produce hallucinations. The proposed architecture and type-aware verification method are validated with HAIMEDA, a real-world medical device damage assessment reporting system developed through Action Design Research. Evaluation shows hallucination detection rates of over 83% for structured entities and 72% for semantic fabrications, with a 30% reduction in report creation time, demonstrating that neuro-symbolic architectures can provide principled safeguards for LLM deployment in data-sensitive domains.

Freie Schlagwort(e)
Neuro-symbolic AI
LLM verification
Formal verification
Trustworthy AI
Hybrid AI
Actor model
AI system architecture
Dokumenttyp
preprint
URI
Datum der Freischaltung
29.05.2026

Permalink: https://fis.uni-bamberg.de/handle/uniba/115305 · Daten als JSON · Zitieren: BibTeX · RIS

Datenstand: 15.11.2026

Eingeschränkter Lesezugriff: Anmeldung, Bearbeitung und Einreichung sind vorübergehend deaktiviert.