arXiv subjects

Florian Matthes

Publications and source records attributed to Florian Matthes.

At least 19 recordsLinked to original sources

arXiv2026

Not All Fallbacks Are Failures: Understanding and Recovering from Fallbacks in Mobile Voice Assistants

Robust understanding of user input is a core requirement for voice assistants deployed in real-world environments. In practice, these systems encounter heterogeneous fallback situations caused by noisy audio input, transcription errors, ambiguous requests, incomplete utterances, or unintended activations. Existing systems typically respond with generic fallback messages, which do not resolve the underlying interaction failure and can degrade user experience. We study fallback handling in a deployed smartwatch-based voice assistant for general health support in everyday environments. Our analysis is based on six months of real-world usage data from more than 500 users, yielding a dataset of 3,030 anonymized, naturally occurring fallback-triggering utterances. We contribute (1) an operational taxonomy and the annotated VoxFallbacks dataset of these interactions, (2) a comparative evaluation of different models within a classification pipeline under practical deployment constraints, and (3) practical lessons for designing robust and cost-efficient fallback mechanisms. Results show that lightweight embedding-based classifiers outperform larger generative models on most classification tasks while requiring substantially fewer computational resources.

cs.CL↗

arXiv2026

PrivBench: A Holistic and Modular Benchmarking Platform for Evaluating Text-to-Text Privatization

Natural Language Processing methods have enabled novel solutions and advances in the field of privacy, particularly in the sub-domain of text-to-text privatization, where the goal is to transform a sensitive input text into a privatized output by ideally masking (in)directly identifiable or otherwise private information. The evaluation of text-to-text privatization, however, is not straightforward, and the extant literature has utilized a myriad of techniques and metrics to quantify the privacy-preserving capabilities of privatization methods. Seeking to unify the evaluation of text-to-text privatization, we introduce PrivBench, a holistic and modular benchmarking platform for researchers and practitioners working on text privatization. PrivBench is holistic in that it evaluates privatization on a series of defined desiderata, which are structured into modules. PrivBench is not only modular but also extensible, allowing for future updates and benchmark versions. PrivBench is user-centered and promotes competition via real-time evaluation and a live public leaderboard. The platform is free to use and openly accessible at https://privbench.com/.

cs.CL↗

arXiv2026

Executable Arbitrage and Market Efficiency in Prediction Markets

Deterministic payoff identities imply no-arbitrage bounds in winner-takes-all prediction markets, but violations of these bounds need not be exploitable before settlement. We distinguish payoff-space no-arbitrage, which follows from terminal payoffs, from protocol-executable no-arbitrage, which depends on the position transformations available to traders. Polymarket's negative-risk markets make this distinction observable: linked binary markets represent mutually exclusive outcomes, while the NegRisk Adapter operationalizes only the NO-to-YES direction before settlement. We reconstruct depth-aware executable portfolio values and combine them with actor-level transaction histories and on-chain conversion traces to measure payoff-bound violations and exploitation. Our reconstruction estimates \$1.12 million in arbitrage profit across two realization channels: \$1.086 million from converter-enabled strategies and \$32 thousand from settlement-based basket formation. In the CLOB sample, positive violations concentrate on the unsupported YES side, whereas adapter-supported NO-side violations are substantially less frequent and shorter-lived. These patterns are consistent with the view that pre-settlement conversion strengthens enforcement by reducing capital lock-up and enabling inventory recycling. Finally, we implement a prototype bidirectional extension of the NegRisk Adapter that makes the reverse path executable before settlement. Together, our findings show that market efficiency depends not only on payoff structure, but also on whether protocols expose payoff equivalences as executable primitives.

cs.CE↗

arXiv2026

sebis at CRF Filling 2026: A Two-Stage Local LLM Pipeline for Medical CRF Filling

The extraction of structured clinical information from unstructured EHR notes is a persistent bottleneck in healthcare informatics. While large language models (LLMs) offer high performance, their deployment in clinical settings is hindered by privacy risks, inference costs, and the tendency to hallucinate beyond textual evidence. We address these challenges for the CL4Health 2026 Case Report Form (CRF) filling task by proposing a fully local, domain-adapted pipeline using the MedGemma-27B model. Our two-stage architecture, which separates binary presence classification from value extraction, enforces strict adherence to textual evidence and ensures deterministic outputs for negated, uncertain, or unknown states. By leveraging item-specific, few-shot in-context learning without external API calls or fine-tuning, our approach achieves a macro-F1 score of 0.55 on the official English test track. This result secures second place among all locally-hosted, open-source submissions. Our work demonstrates that privacy-preserving, on-premise LLM pipelines can achieve near-competitive performance with proprietary frontier models, providing a practical, data-sovereign framework for clinical NLP.

cs.CL↗

arXiv2026

When Certainty Is Not Worth It: Capital Lock-Up and Settlement Discounting in Prediction Markets

Collateralized prediction markets are contingent-claim markets in which economic uncertainty can disappear before winning claims become redeemable. This paper studies the pricing effect of that delay. When collateral remains locked until oracle settlement, a near-certain dollar is a delayed dollar, so prices embed a maturity-dependent settlement discount in addition to beliefs about outcomes. We recover an implied settlement-discount term structure from persistent near-certain contracts using realized settlement times and summarize it as an annualized settlement wedge (ASW). The recovered wedges are positive, maturity-dependent, and time-varying. Adjusting pricesby these curves reduces the near-certainty horizon gradient by roughly 48-88%, indicating that much of the raw maturity pattern reflects priced settlement frictions rather than forecast error alone. Market architecture changes the wedge: negRisk conversion compresses discounts by recycling part of the position into synthetic collateral, while yield-bearing collateral flattens the term structure by reducing the opportunity cost of lock-up. The results show that pricing quality in prediction markets is endogenous to settlement mechanics, collateral productivity, and capital-recycling design. Prediction-market prices therefore aggregate information through a financial infrastructure whose funding conditions are measurable and economically important.

cs.CE↗

arXiv2026

Atomic Fact-Checking Increases Clinician Trust in Large Language Model Recommendations for Oncology Decision Support: A Randomized Controlled Trial

Question: Does atomic fact-checking, which decomposes AI treatment recommendations into individually verifiable claims linked to source guideline documents, increase clinician trust compared to traditional explainability approaches? Findings: In this randomized trial of 356 clinicians generating 7,476 trust ratings, atomic fact-checking produced a large effect on trust (Cohen's d = 0.94), increasing the proportion of clinicians expressing trust from 26.9% to 66.5%. Traditional transparency mechanisms showed a dose-response gradient of improvement over baseline (d = 0.25 to 0.50). Meaning: Decomposing AI recommendations into individually verifiable claims linked to source guidelines produces substantially higher clinician trust than traditional explainability approaches in high-stakes clinical decisions.

cs.CL↗

arXiv2026

A Systematic Exploration of Text Decomposition and Budget Distribution in Differentially Private Text Obfuscation

The goal of differentially private text obfuscation is to obfuscate, or "perturb", input texts with Differential Privacy (DP) guarantees, such that the private output texts are quantifiably indistinguishable from the originals. While perturbation at the word level is intuitive, meaningful text privatization happens on complete documents. Recent research has laid the groundwork for reasoning about privacy budget distribution, namely, how an overall $\varepsilon$ budget can be sensibly distributed among the component pieces of a text. We perform a systematic evaluation of multiple text decomposition and budget distribution techniques in the context of DP text obfuscation, testing how different methods for chunking texts can be combined with techniques for allocating $\varepsilon$ to these chunks. Our experiments reveal that such design choices are very important, as even with comparable privacy budgets, significantly different results can occur based on which methods are chosen. In this, we provide credible evidence of the feasibility of maximizing empirical trade-offs by optimizing DP obfuscation procedures.

cs.CL↗

arXiv2026

A Case Study on the Impact of Anonymization Along the RAG Pipeline

Despite the considerable promise of Retrieval-Augmented Generation (RAG), many real-world use cases may create privacy concerns, where the purported utility of RAG-enabled insights comes at the risk of exposing private information to either the LLM or the end user requesting the response. As a potential mitigation, using anonymization techniques to remove personally identifiable information (PII) and other sensitive markers in the underlying data represents a practical and sensible course of action for RAG administrators. Despite a wealth of literature on the topic, no works consider the placement of anonymization along the RAG pipeline, i.e., asking the question, where should anonymization happen? In this case study, we systematically and empirically measure the impact of anonymization at two important points along the RAG pipeline: the dataset and generated answer. We show that differences in privacy-utility trade-offs can be observed depending on where anonymization took place, demonstrating the significance of privacy risk mitigation placement in RAG.

cs.CR↗

arXiv2026

sebis at ArchEHR-QA 2026: How Much Can You Do Locally? Evaluating Grounded EHR QA on a Single Notebook

Clinical question answering over electronic health records (EHRs) can help clinicians and patients access relevant medical information more efficiently. However, many recent approaches rely on large cloud-based models, which are difficult to deploy in clinical environments due to privacy constraints and computational requirements. In this work, we investigate how far grounded EHR question answering can be pushed when restricted to a single notebook. We participate in all four subtasks of the ArchEHR-QA 2026 shared task and evaluate several approaches designed to run on commodity hardware. All experiments are conducted locally without external APIs or cloud infrastructure. Our results show that such systems can achieve competitive performance on the shared task leaderboards. In particular, our submissions perform above average in two subtasks, and we observe that smaller models can approach the performance of much larger systems when properly configured. These findings suggest that privacy-preserving EHR QA systems running fully locally are feasible with current models and commodity hardware. The source code is available at https://github.com/ibrahimey/ArchEHR-QA-2026.

cs.CL↗

arXiv2026

Real-Time Generation of Game Video Commentary with Multimodal LLMs: Pause-Aware Decoding Approaches

Real-time video commentary generation provides textual descriptions of ongoing events in videos. It supports accessibility and engagement in domains such as sports, esports, and livestreaming. Commentary generation involves two essential decisions: what to say and when to say it. While recent prompting-based approaches using multimodal large language models (MLLMs) have shown strong performance in content generation, they largely ignore the timing aspect. We investigate whether in-context prompting alone can support real-time commentary generation that is both semantically relevant and well-timed. We propose two prompting-based decoding strategies: 1) a fixed-interval approach, and 2) a novel dynamic interval-based decoding approach that adjusts the next prediction timing based on the estimated duration of the previous utterance. Both methods enable pause-aware generation without any fine-tuning. Experiments on Japanese and English datasets of racing and fighting games show that the dynamic interval-based decoding can generate commentary more closely aligned with human utterance timing and content using prompting alone. We release a multilingual benchmark dataset, trained models, and implementations to support future research on real-time video commentary generation.

cs.CL↗

arXiv2026

Privacy Starts with UI: Privacy Patterns and Designer Perspectives in UI/UX Practice

In the study of Human-Computer Interaction, privacy is often seen as a core issue, and it has been explored directly in connection with User Interface (UI) and User Experience (UX) design. We systematically investigate the key considerations and factors for privacy in UI/UX, drawing upon the extant literature and 15 semi-structured interviews with experts working in the field. These insights lead to the synthesis of 14 primary design considerations for privacy in UI/UX, as well as 14 key factors under four main axes affecting privacy work therein. From these findings, we produce our main research artifact, a UI/UX Privacy Pattern Catalog, which we validate in a series of two interactive workshops and one online survey with UI/UX practitioners. Our work not only systematizes a field growing in both attention and importance, but it also provides an actionable and expert-validated artifact to guide UI/UX designers in realizing privacy-preserving UI/UX design.

cs.HC↗

arXiv2026

SoK: Privacy Risks and Mitigations in Retrieval-Augmented Generation Systems

The continued promise of Large Language Models (LLMs), particularly in their natural language understanding and generation capabilities, has driven a rapidly increasing interest in identifying and developing LLM use cases. In an effort to complement the ingrained "knowledge" of LLMs, Retrieval-Augmented Generation (RAG) techniques have become widely popular. At its core, RAG involves the coupling of LLMs with domain-specific knowledge bases, whereby the generation of a response to a user question is augmented with contextual and up-to-date information. The proliferation of RAG has sparked concerns about data privacy, particularly with the inherent risks that arise when leveraging databases with potentially sensitive information. Numerous recent works have explored various aspects of privacy risks in RAG systems, from adversarial attacks to proposed mitigations. With the goal of surveying and unifying these works, we ask one simple question: What are the privacy risks in RAG, and how can they be measured and mitigated? To answer this question, we conduct a systematic literature review of RAG works addressing privacy, and we systematize our findings into a comprehensive set of privacy risks, mitigation techniques, and evaluation strategies. We supplement these findings with two primary artifacts: a Taxonomy of RAG Privacy Risks and a RAG Privacy Process Diagram. Our work contributes to the study of privacy in RAG not only by conducting the first systematization of risks and mitigations, but also by uncovering important considerations when mitigating privacy risks in RAG systems and assessing the current maturity of proposed mitigations.

cs.CR↗

arXiv2026

Semantic Non-Fungibility and Violations of the Law of One Price in Prediction Markets

Prediction markets are designed to aggregate dispersed information about future events, yet today's ecosystem is fragmented across heterogeneous operator-run platforms and blockchain-based protocols that independently list economically identical events. In the absence of a shared notion of event identity, liquidity fails to pool across venues, arbitrage becomes capital-intensive or unenforceable, and prices systematically violate the Law of One Price. As a result, market prices reflect platform-local beliefs rather than a single, globally aggregated probability, undermining the core information-aggregation function of prediction markets. We address this gap by introducing a semantic alignment framework that makes cross-platform event identity explicit through joint analysis of natural-language descriptions, resolution semantics, and temporal scope. Applying this framework, we construct the first human-validated, cross-platform dataset of aligned prediction markets, covering over 100 000 events across ten major venues from 2018 to 2025. Using this dataset, we show that roughly 6% of all events are concurrently listed across platforms and that semantically equivalent markets exhibit persistent execution-aware price deviations of 2-4% on average, even in highly liquid and information-rich settings. These mispricings give rise to persistent cross-platform arbitrage opportunities driven by structural frictions rather than informational disagreement. Overall, our results demonstrate that semantic non-fungibility is a fundamental barrier to price convergence, and that resolving event identity is a prerequisite for prediction markets to aggregate information at a global scale.

cs.CE↗

arXiv2025

Improving Reliability and Explainability of Medical Question Answering through Atomic Fact Checking in Retrieval-Augmented LLMs

Large language models (LLMs) exhibit extensive medical knowledge but are prone to hallucinations and inaccurate citations, which pose a challenge to their clinical adoption and regulatory compliance. Current methods, such as Retrieval Augmented Generation, partially address these issues by grounding answers in source documents, but hallucinations and low fact-level explainability persist. In this work, we introduce a novel atomic fact-checking framework designed to enhance the reliability and explainability of LLMs used in medical long-form question answering. This method decomposes LLM-generated responses into discrete, verifiable units called atomic facts, each of which is independently verified against an authoritative knowledge base of medical guidelines. This approach enables targeted correction of errors and direct tracing to source literature, thereby improving the factual accuracy and explainability of medical Q&A. Extensive evaluation using multi-reader assessments by medical experts and an automated open Q&A benchmark demonstrated significant improvements in factual accuracy and explainability. Our framework achieved up to a 40% overall answer improvement and a 50% hallucination detection rate. The ability to trace each atomic fact back to the most relevant chunks from the database provides a granular, transparent explanation of the generated responses, addressing a major gap in current medical AI applications. This work represents a crucial step towards more trustworthy and reliable clinical applications of LLMs, addressing key prerequisites for clinical application and fostering greater confidence in AI-assisted healthcare.

cs.CL↗

arXiv2025

With Privacy, Size Matters: On the Importance of Dataset Size in Differentially Private Text Rewriting

Recent work in Differential Privacy with Natural Language Processing (DP NLP) has proposed numerous promising techniques in the form of text rewriting mechanisms. In the evaluation of these mechanisms, an often-ignored aspect is that of dataset size, or rather, the effect of dataset size on a mechanism's efficacy for utility and privacy preservation. In this work, we are the first to introduce this factor in the evaluation of DP text privatization, where we design utility and privacy tests on large-scale datasets with dynamic split sizes. We run these tests on datasets of varying size with up to one million texts, and we focus on quantifying the effect of increasing dataset size on the privacy-utility trade-off. Our findings reveal that dataset size plays an integral part in evaluating DP text rewriting mechanisms; additionally, these findings call for more rigorous evaluation procedures in DP NLP, as well as shed light on the future of DP NLP in practice and at scale.

cs.CL↗

arXiv2025

Cross-Chain Sealed-Bid Auctions Using Confidential Compute Blockchains

Sealed-bid auctions ensure fair competition and efficient allocation but are often deployed on centralized infrastructure, enabling opaque manipulation. Public blockchains eliminate central control, yet their inherent transparency conflicts with the confidentiality required for sealed bidding. Prior attempts struggle to reconcile privacy, verifiability, and scalability without relying on trusted intermediaries, multi-round protocols, or expensive cryptography. We present a sealed-bid auction protocol that executes sensitive bidding logic on a Trusted Execution Environment (TEE)-backed confidential compute blockchain while retaining settlement and enforcement on a public chain. Bidders commit funds to enclave-generated escrow addresses, ensuring confidentiality and binding commitments. After the deadline, any party can trigger resolution: the confidential blockchain determines the winner through verifiable off-chain computation and issues signed settlement transactions for execution on the public chain. Our design provides security, privacy, and scalability without trusted third parties or protocol modifications. We implement it on SUAVE with Ethereum settlement, evaluate its scalability and trust assumptions, and demonstrate deployment with minimal integration on existing infrastructure

cs.CR↗

arXiv2025

"We are not Future-ready": Understanding AI Privacy Risks and Existing Mitigation Strategies from the Perspective of AI Developers in Europe

The proliferation of AI has sparked privacy concerns related to training data, model interfaces, downstream applications, and more. We interviewed 25 AI developers based in Europe to understand which privacy threats they believe pose the greatest risk to users, developers, and businesses and what protective strategies, if any, would help to mitigate them. We find that there is little consensus among AI developers on the relative ranking of privacy risks. These differences stem from salient reasoning patterns that often relate to human rather than purely technical factors. Furthermore, while AI developers are aware of proposed mitigation strategies for addressing these risks, they reported minimal real-world adoption. Our findings highlight both gaps and opportunities for empowering AI developers to better address privacy risks in AI.

cs.HC↗

arXiv2025

CRSet: Private Non-Interactive Verifiable Credential Revocation

Like any digital certificate, Verifiable Credentials (VCs) require a way to revoke them in case of an error or key compromise. Existing solutions for VC revocation, most prominently Bitstring Status List, are not viable for many use cases because they may leak the issuer's activity, which in turn leaks internal business metrics. For instance, staff fluctuation through the revocation of employee IDs. We identify the protection of issuer activity as a key gap and propose a formal definition for a corresponding characteristic of a revocation mechanism. Then, we introduce CRSet, a non-interactive mechanism that trades some space efficiency to reach these privacy characteristics. For that, we provide a proof sketch. Issuers periodically encode revocation data and publish it via Ethereum blob-carrying transactions, ensuring secure and private availability. Relying Parties (RPs) can download it to perform revocation checks locally. Sticking to a non-interactive design also makes adoption easier because it requires no changes to wallet agents and exchange protocols. We also implement and empirically evaluate CRSet, finding its real-world behavior to match expectations. One Ethereum blob fits revocation data for about 170k VCs.

cs.CR↗