Florian Matthes
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Florian Matthes has a total of 182 co-authors. Here are the top co-authors by number of shared publications:
- Juraj Vladika
Technical University of Munich
27 shared publications - Stephen Meisenbacher
Technical University of Munich
26 shared publications - Anum Afzal
Technical University of Munich Technical University of Munich Nvidia Research & Bar-Ilan University Nvidia Research
12 shared publications - Alexandra Klymenko
Technical University of Munich, Garching, Germany
9 shared publications - Daniel Braun
Department of High-Tech Business and Entrepreneurship, University of Twente, Enschede, The Netherlands
8 shared publications - Franziska Tobisch
Technical University of Munich
8 shared publications
Latest Publications
- Open Access
- 13 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 1835-1842
Auditing Automated ESG Disclosure Measurement from Web Archives
- 13 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 1835-1842
Open AccessAuditing Automated ESG Disclosure Measurement from Web Archives
- Open Access
- 21 September 2026
- Published by JMIR Publications Inc. in Journal of Medical Internet Research
- Vol. 28, e92090
Improving Reliability and Explainability of Medical Question Answering Through Atomic Fact-Checking in Retrieval-Augmented Large Language Models: Creation and Validation Study
Background: Large language models (LLMs) exhibit extensive medical knowledge but are prone to hallucinations and show low fact-level explainability, limiting clinical adoption and regulatory compliance. Existing approaches, such as retrieval-augmented generation, partially address these issues by grounding answers in source documents; however, the aforementioned problems persist. Objective: We propose the application of an atomic fact-checking framework designed to enhance the reliability and explainability of LLMs in medical long-form question answering. By decomposing generated answers into discrete atomic facts and verifying each against an authoritative knowledge base of medical guidelines, this approach enables precise identification and correction of incorrect statements, alongside explicit linkage to supporting literature. Methods: The fact-checking algorithm operates within a retrieval-augmented generation framework: LLM-generated answers are decomposed into atomic facts (smallest and self-contained information units), each of which is assessed and corrected if FALSE. To determine an optimal strategy, the validation–question and answer (Q&A) set on prostate cancer treatment was tested under varying instructions. An extensive evaluation, including multireader assessments by human medical experts and the automated open Q&A benchmark AMEGA (Autonomous Medical Evaluation for Guideline Adherence), was conducted for the final pipeline. In addition to another radiooncologic test–Q&A set, anonymized real-world tumor board cases and an independent, established neurology-Q&A set were used. Given their transparency and accessibility advantages, we compared various open-source models in pairs of generalist models and their medical fine-tuned counterparts, with regard to performance and improvements by fact-checking. Results: The framework significantly reduced hallucinations and inaccuracies. Medical expert assessment and automated benchmarks demonstrated significant improvements in factual accuracy, achieving up to a 50% overall answer improvement and an 80% hallucination detection rate. Notably, the observed gain was strongest in real tumor-board questions—the most challenging dataset. Additionally, the framework achieved high explainability by tracing each atomic fact back to the most relevant chunks from the database, providing a granular, transparent explanation of the generated responses. Conclusions: To conclude, we present the application of an atomic fact-checking algorithm to medical Q&A. It identifies factual inaccuracies and hallucinations in LLM-generated answers, achieving the greatest gains on clinically realistic, complex questions. Correction via fact-checking improves the overall answer quality while achieving fact-wise explainability, paving the way for more credible clinical use of LLMs.
- 21 September 2026
- Published by JMIR Publications Inc. in Journal of Medical Internet Research
- Vol. 28, e92090
Open AccessImproving Reliability and Explainability of Medical Question Answering Through Atomic Fact-Checking in Retrieval-Augmented Large Language Models: Creation and Validation Study
Background: Large language models (LLMs) exhibit extensive medical knowledge but are prone to hallucinations and show low fact-level explainability, limiting clinical adoption and regulatory compliance. Existing approaches, such as retrieval-augmented generation, partially address these issues by grounding answers in source documents; however, the aforementioned problems persist. Objective: We propose the application of an atomic fact-checking framework designed to enhance the reliability and explainability of LLMs in medical long-form question answering. By decomposing generated answers into discrete atomic facts and verifying each against an authoritative knowledge base of medical guidelines, this approach enables precise identification and correction of incorrect statements, alongside explicit linkage to supporting literature. Methods: The fact-checking algorithm operates within a retrieval-augmented generation framework: LLM-generated answers are decomposed into atomic facts (smallest and self-contained information units), each of which is assessed and corrected if FALSE. To determine an optimal strategy, the validation–question and answer (Q&A) set on prostate cancer treatment was tested under varying instructions. An extensive evaluation, including multireader assessments by human medical experts and the automated open Q&A benchmark AMEGA (Autonomous Medical Evaluation for Guideline Adherence), was conducted for the final pipeline. In addition to another radiooncologic test–Q&A set, anonymized real-world tumor board cases and an independent, established neurology-Q&A set were used. Given their transparency and accessibility advantages, we compared various open-source models in pairs of generalist models and their medical fine-tuned counterparts, with regard to performance and improvements by fact-checking. Results: The framework significantly reduced hallucinations and inaccuracies. Medical expert assessment and automated benchmarks demonstrated significant improvements in factual accuracy, achieving up to a 50% overall answer improvement and an 80% hallucination detection rate. Notably, the observed gain was strongest in real tumor-board questions—the most challenging dataset. Additionally, the framework achieved high explainability by tracing each atomic fact back to the most relevant chunks from the database, providing a granular, transparent explanation of the generated responses. Conclusions: To conclude, we present the application of an atomic fact-checking algorithm to medical Q&A. It identifies factual inaccuracies and hallucinations in LLM-generated answers, achieving the greatest gains on clinically realistic, complex questions. Correction via fact-checking improves the overall answer quality while achieving fact-wise explainability, paving the way for more credible clinical use of LLMs.
- 24 August 2026
- Published by Springer Nature
- p. 81-98
Improving the Utility, Privacy, and Human Acceptance of Differentially Private Rewritten Texts
- 24 August 2026
- Published by Springer Nature
- p. 81-98
Improving the Utility, Privacy, and Human Acceptance of Differentially Private Rewritten Texts
- Open Access
- 20 July 2026
- Published by Association for Computing Machinery (ACM) in ACM SIGMETRICS Performance Evaluation Review
- Vol. 54 (1) , 19-21
Cross-Chain Arbitrage: The Next Frontier of MEV in Decentralized Finance
Decentralized finance (DeFi) markets are fragmented across block-chains, creating price differences that arbitrage helps eliminate. Today, much of this price alignment occurs against centralized exchanges, whose deep liquidity and fast execution make them dominant venues for price discovery. As trading activity moves on-chain, cross-chain arbitrage between decentralized exchanges is becoming increasingly important. Yet despite its relevance to DeFi, prior work has largely been limited to hypothetical opportunity analyses and conceptual overviews. We conduct a year-long empirical study of cross-chain arbitrage from September 2023 to August 2024 across nine blockchains and identify 242,535 executed arbitrages totaling USD 868.64 million in volume. Activity grows by 5.5x over the study period and surges after the Dencun upgrade on March 13, 2024. Most trades rely on pre-positioned inventory (66.96%) and settle in a median of 9s, whereas bridge-based arbitrages take 242s, highlighting the latency cost of today's bridges. Market concentration is substantial: the five largest addresses execute more than half of all trades, and one address alone captures nearly 40% of daily volume post-Dencun. Taken together, these patterns suggest that cross-chain arbitrage can encourage vertical integration, concentrating sequencing infrastructure and economic power, thereby increasing censorship and consensus security risks. Decentralizing block building and lowering entry barriers are therefore important mitigation directions.
- 20 July 2026
- Published by Association for Computing Machinery (ACM) in ACM SIGMETRICS Performance Evaluation Review
- Vol. 54 (1) , 19-21
Open AccessCross-Chain Arbitrage: The Next Frontier of MEV in Decentralized Finance
Decentralized finance (DeFi) markets are fragmented across block-chains, creating price differences that arbitrage helps eliminate. Today, much of this price alignment occurs against centralized exchanges, whose deep liquidity and fast execution make them dominant venues for price discovery. As trading activity moves on-chain, cross-chain arbitrage between decentralized exchanges is becoming increasingly important. Yet despite its relevance to DeFi, prior work has largely been limited to hypothetical opportunity analyses and conceptual overviews. We conduct a year-long empirical study of cross-chain arbitrage from September 2023 to August 2024 across nine blockchains and identify 242,535 executed arbitrages totaling USD 868.64 million in volume. Activity grows by 5.5x over the study period and surges after the Dencun upgrade on March 13, 2024. Most trades rely on pre-positioned inventory (66.96%) and settle in a median of 9s, whereas bridge-based arbitrages take 242s, highlighting the latency cost of today's bridges. Market concentration is substantial: the five largest addresses execute more than half of all trades, and one address alone captures nearly 40% of daily volume post-Dencun. Taken together, these patterns suggest that cross-chain arbitrage can encourage vertical integration, concentrating sequencing infrastructure and economic power, thereby increasing censorship and consensus security risks. Decentralizing block building and lowering entry barriers are therefore important mitigation directions.
- 2 July 2026
- Published by Springer Nature
- p. 627-641
Dynamic User Assistance for Component-Based Web Interfaces Using Large Language Models
- 2 July 2026
- Published by Springer Nature
- p. 627-641
Dynamic User Assistance for Component-Based Web Interfaces Using Large Language Models
- 2 July 2026
- Published by Springer Nature
- p. 858-871
Enhancing Conversational Access to Patient Health Data Through LLM-Powered Chat Interfaces
- 2 July 2026
- Published by Springer Nature
- p. 858-871
Enhancing Conversational Access to Patient Health Data Through LLM-Powered Chat Interfaces
- Open Access
- 22 June 2026
- Published by Association for Computing Machinery (ACM)
- p. 3-9
A Case Study on the Impact of Anonymization Along the RAG Pipeline
- 22 June 2026
- Published by Association for Computing Machinery (ACM)
- p. 3-9
Open AccessA Case Study on the Impact of Anonymization Along the RAG Pipeline
- Open Access
- 8 June 2026
- Published by Association for Computing Machinery (ACM)
- p. 19-21
Cross-Chain Arbitrage: The Next Frontier of MEV in Decentralized Finance
- 8 June 2026
- Published by Association for Computing Machinery (ACM)
- p. 19-21
Open AccessCross-Chain Arbitrage: The Next Frontier of MEV in Decentralized Finance
- 26 May 2026
- Published by Springer Nature
- p. 45-59
Guiding Principles for Government as a Platform
- Open Access
- 23 March 2026
- Published by Association for Computing Machinery (ACM)
- p. 1010-1017
Supporting Agility at Scale Through an Organization-wide Software Engineering Community of Practice: Insights from a Large Software Development Organization
- 23 March 2026
- Published by Association for Computing Machinery (ACM)
- p. 1010-1017
Open AccessSupporting Agility at Scale Through an Organization-wide Software Engineering Community of Practice: Insights from a Large Software Development Organization