https://arxiv.org/api/X52n6LMmvtj3mUmXIYP80GWAkAQ 2026-09-25T17:38:57Z 12 23 0 http://arxiv.org/abs/2608.22025v1 2026-08-22T16:03:36Z

Trust and reputation systems underpin reliable interactions in large, distributed networks. However, conventional models typically propagate trust only forward, offering no accountability for endorsers regarding whom they vouch for, and leaving newly joined nodes without a meaningful initial reputation. RepuLink addresses these limitations by proposing a two-layer trust and reputation model that integrates direct interaction feedback with domain-specific endorsements. Crucially, it holds endorsers accountable via Backward Endorsement Penalty/Reward Propagation (BEPP/BERP). This paper demonstrates RepuLink-Tool, a deployable, full-stack reference implementation of this model. The application enables nodes to interact, rate, and endorse each other, while tracking reputation via a live dashboard and an interactive trust network graph. Furthermore, we introduce a new Linked Data layer built on top of the application. This layer features a lightweight OWL ontology encompassing nodes, interactions, ratings, endorsements, pairwise trust assessments, and computed reputation scores annotated with PROV-O provenance. It also provides an on-the-fly RDF projection of each user's trust network in multiple serialisations, alongside a scoped SPARQL endpoint that nodes can query live against their own data.

2026-08-22T16:03:36Z ISWC 2026: 25th International Semantic Web Conference, Poster and Demos Wenbo Wu George Konstantinidis http://arxiv.org/abs/2607.07109v1 2026-07-08T07:50:49Z

The EU Cyber Resilience Act (CRA) makes a smart bet. It does not demand that products be free of vulnerabilities, but only that manufacturers run a process: assess risk, handle flaws, ship updates. The bet pays off if four things about the world stay true: (P1) finding vulnerabilities is slow, skilled, human work; (P2) a product's exploitable flaws are knowable the day it ships; (P3) exploitation is rare enough to notice; and (P4) fixes keep pace with discovery. Cybersecurity AI (CAI) agents, AI put to work finding and exploiting flaws in other products, falsify all four. The regime answers in two opposite ways. Against the sheer volume of flaws that agents surface it bends (P1): built for scarce attention, it re-centres compliance on defensible, documented prioritisation, and holds. But agents also collapse the speed and economics of the vulnerability lifecycle, and here it breaks (P2, P3, P4): a product that passed every check becomes exploitable without anyone touching it, so its market-entry test, its reporting trigger, and its one-and-done certificate vouch for a security that has quietly expired. The fault is in the landscape, not the product, so running the process more diligently cannot repair it. We map each mechanism to the force that strains or snaps it, and find the cure and the disease cut from the same cloth: because defenders and attackers wield the same AI, the only conformity that survives is one that never stops running. We also carry the remedy from proposal to proof on two CRA-scope robots, a humanoid and a lawn mower, where an agentic defender holds a line their undefended selves cannot. On the evidence already in hand, the CRA reaches full force in December 2027 certifying products against a world that has already changed. Static, human-paced security is finished; what replaces it must be continuous and agent-operated, and that is no longer a matter of taste.

2026-07-08T07:50:49Z Víctor Mayoral-Vilches http://arxiv.org/abs/2602.06172v1 2026-02-05T20:24:54Z

Biological AI tools for protein design and structure prediction are advancing rapidly, creating dual-use risks that existing safeguards cannot adequately address. Current model-level restrictions, including keyword filtering, output screening, and content-based access denials, are fundamentally ill-suited to biology, where reliable function prediction remains beyond reach and novel threats evade detection by design. We propose a three-tier Know Your Customer (KYC) framework, inspired by anti-money laundering (AML) practices in the financial sector, that shifts governance from content inspection to user verification and monitoring. Tier I leverages research institutions as trust anchors to vouch for affiliated researchers and assume responsibility for vetting. Tier II applies output screening through sequence homology searches and functional annotation. Tier III monitors behavioral patterns to detect anomalies inconsistent with declared research purposes. This layered approach preserves access for legitimate researchers while raising the cost of misuse through institutional accountability and traceability. The framework can be implemented immediately using existing institutional infrastructure, requiring no new legislation or regulatory mandates.

2026-02-05T20:24:54Z Jonathan Feldman Tal Feldman Annie I Anton http://arxiv.org/abs/2509.05390v1 2025-09-05T10:11:56Z

The use of large language models (LLMs) in bioethical, scientific, and medical writing remains controversial. While there is broad agreement in some circles that LLMs cannot count as authors, there is no consensus about whether and how humans using LLMs can count as authors. In many fields, authorship is distributed among large teams of researchers, some of whom, including paradigmatic senior authors who guide and determine the scope of a project and ultimately vouch for its integrity, may not write a single word. In this paper, we argue that LLM use (under specific conditions) is analogous to a form of senior authorship. On this view, the use of LLMs, even to generate complete drafts of research papers, can be considered a legitimate form of authorship according to the accepted criteria in many fields. We conclude that either such use should be recognized as legitimate, or current criteria for authorship require fundamental revision. AI use declaration: GPT-5 was used to help format Box 1. AI was not used for any other part of the preparation or writing of this manuscript.

2025-09-05T10:11:56Z 28 pages, 0 figures Clint Hurshman Sebastian Porsdam Mann Julian Savulescu Brian D. Earp http://arxiv.org/abs/2404.05088v1 2024-04-07T22:06:19Z

In this paper, we assess the robustness (reliability) of ChatGPT under input perturbations for one of the most fundamental tasks of Information Extraction (IE) i.e. Named Entity Recognition (NER). Despite the hype, the majority of the researchers have vouched for its language understanding and generation capabilities; a little attention has been paid to understand its robustness: How the input-perturbations affect 1) the predictions, 2) the confidence of predictions and 3) the quality of rationale behind its prediction. We perform a systematic analysis of ChatGPT's robustness (under both zero-shot and few-shot setup) on two NER datasets using both automatic and human evaluation. Based on automatic evaluation metrics, we find that 1) ChatGPT is more brittle on Drug or Disease replacements (rare entities) compared to the perturbations on widely known Person or Location entities, 2) the quality of explanations for the same entity considerably differ under different types of "Entity-Specific" and "Context-Specific" perturbations and the quality can be significantly improved using in-context learning, and 3) it is overconfident for majority of the incorrect predictions, and hence it could lead to misguidance of the end-users.

2024-04-07T22:06:19Z 3 Figures, 7 Tables Ishani Mondal Abhilasha Sancheti http://arxiv.org/abs/2212.12420v2 2023-03-07T08:31:18Z

Multi-link Operation (MLO), arguably the most disruptive feature introduced in IEEE 802.11be, will cater for delay-sensitive applications by using multiple radio interfaces concurrently. In this paper, we analyze the delay distribution of MLO under non-saturated traffic. Our results show that upgrading from legacy single-link to MLO successfully contains the $95$th percentile delay as the traffic load increases, but also that adding extra interfaces yields rapidly diminishing gains. Further experiments we conduct on Google Stadia traces, accounting for realistic traffic with batch arrivals, confirm the insights stemming from our theoretical analysis, vouching for MLO to support high-resolution real-time video with a controlled delay.

2022-12-23T15:58:31Z IEEE Wireless Communications Letters, 2023 Boris Bellalta Marc Carrascosa Lorenzo Galati-Giordano Giovanni Geraci 10.1109/LWC.2023.3235001 http://arxiv.org/abs/2212.06436v1 2022-12-13T08:57:42Z

Nowadays, most business and social interactions have moved to the internet, highlighting the relevance of creating online trust. One way to obtain a measure of trust is through reputation mechanisms, which record one's past performance and interactions to generate a reputational value. We observe that numerous existing reputation mechanisms share similarities with actual social phenomena; we call such mechanisms 'social reputation mechanisms'. The aim of this paper is to discuss several social phenomena and map these to existing social reputation mechanisms in a variety of scopes. First, we focus on reputation mechanisms in the individual scope, in which everyone is responsible for their own reputation. Subjective reputational values may be communicated to different entities in the form of recommendations. Secondly, we discuss social reputation mechanisms in the acquaintances scope, where one's reputation can be tied to another through vouching or invite-only networks. Finally, we present existing social reputation mechanisms in the neighbourhood scope. In such systems, one's reputation can heavily be affected by the behaviour of others in their neighbourhood or social group.

2022-12-13T08:57:42Z 10 pages, 3 figures, 1 table Thomas Werthenbach Johan Pouwelse http://arxiv.org/abs/2211.01179v3 2024-09-25T20:19:42Z

This paper presents Solidago, an end-to-end modular pipeline to allow any community of users to collaboratively score any number of entities. Solidago proposes a six-module decomposition. First, it uses pretrust and peer-to-peer vouches to assign trust scores to users. Second, based on participation, trust scores are turned into voting rights per user per entity. Third, for each user, a preference model is learned from the user's evaluation data. Fourth, users' models are put on a similar scale. Fifth, these models are securely aggregated. Sixth, models are post-processed to yield human-readable global scores. We also propose default implementations of the six modules, including a novel trust propagation algorithm, and adaptations of state-of-the-art scaling and aggregation solutions. Our pipeline has been successfully deployed on the open-source platform tournesol.app. We thereby lay an appealing foundation for the collaborative, effective, scalable, fair, interpretable and secure scoring of any set of entities.

2022-10-30T09:27:16Z 34 pages, 10 figures Lê Nguyên Hoang Romain Beylerian Bérangère Colbois Julien Fageot Louis Faucon Aidan Jungo Alain Le Noac'h Adrien Matissart Oscar Villemaud http://arxiv.org/abs/2210.10505v2 2023-03-13T08:25:30Z

Typically, in quiescent conditions, attractive colloids at low volume fractions form fractal gels structured into two length scales: the colloidal and the fractal cluster scales. However when flow interfere with gelation colloidal fractal gels may display three distinct length scales [Dagès, et al., Soft Matter 18, 6645 (2022)]. Following those recent experimental investigations, we derive two models that account for the structure and the rheological properties of such atypical colloidal gels. The gel elasticity is inferred from scaling arguments and the structure is translated into scattering intensities following the global scattering functions approach proposed by Beaucage and typically measured in small angle X-ray scattering (SAXS). In both models, we consider that the colloids condensate into fractal clusters. In the clusters of clusters model, the clusters form superagregates which then build the gel network. In the interpenetrating clusters model, the clusters interpenetrate one-another to form the gel network. Those two models are then utilised to analyse rheo-SAXS experiments carried out on carbon black gels formed through flow cessation. The results of the analysis vouch for the clusters of clusters model with a densification of the structures as the gel characteristic length scales increase.

2022-10-19T12:21:30Z Journal of Rheology 67, 621 (2023) Louis-Vincent Bouthier Thomas Gibaud 10.1122/8.0000595 http://arxiv.org/abs/2208.04624v1 2022-08-09T09:39:46Z

Self sovereign identity is a form of decentralised credential management. During credential verification, data exchange only happens between the data owner and the verifier without passing through any third parties. While this approach offers a privacy-centric solution, it poses a challenge. How do verifiers trust that the credential is vouched by a trusted source? More specifically, how do verifiers know that the issuer has the reputation or is authorised to issue the credential? In this paper, we propose a trust registry design that handles the aspect of human trust in self sovereign identity. We also introduce an incentivisation mechanism for the trust registry in order to motivate each stakeholder to participate actively and honestly.

2022-08-09T09:39:46Z 10 pages, published in Blockchain Congress Conference 2022 Kai Jun Eer Jesus Diaz Markulf Kohlweiss http://arxiv.org/abs/2106.15688v1 2021-06-29T19:40:47Z

Ditopic bis-(triazole-pyridine)viologens are bidentate ligands that self-assemble into coordination polymers. In such photo-responsive materials, light irradiation initiates photo-induced electron transfer to generate pi-radicals that can self-associate to form pi-dimers. This leads to a cascade of events: processes at the supramolecular scale associated with mechanical and structural transition at the macroscopic scale. By tuning the irradiation power and duration, we evidence the formation of aggregates and gels. Using microscopy, we show that the aggregates are dense polydisperse micron size spindle shaped particles which grow in time. Using microscopy and time resolved micro-rheology, we follow the gelation kinetics which leads to a gel characterized by a correlation length of a few microns and a weak elastic modulus. The analysis of the aggregates and the gel states vouch for an arrested phase separation process.

2021-06-29T19:40:47Z Shagor Chowdhury Quentin Reynard-Feytis Clément Roizard Denis Frath Floris Chevallier Christophe Bucher Thomas Gibaud http://arxiv.org/abs/2009.12150v1 2020-09-25T12:02:08Z

Background. The current cognitive state, such as cognitive effort and depletion, incidental affect or stress may impact the strength of a chosen password unconsciously. Aim. We investigate the effect of incidental fear and stress on the measured strength of a chosen password. Method. We conducted two experiments with within-subject designs measuring the Zxcvbn \textsf{log10} number of guesses as strength of chosen passwords as dependent variable. In both experiments, participants were signed up to a site holding their personal data and, for the second run a day later, asked under a security incident pretext to change their password. (a) Fear. $N_\mathsf{F} = 34$ participants were exposed to standardized fear and happiness stimulus videos in random order. (b) \textbf{Stress.} $N_\mathsf{S} = 50$ participants were either exposed to a battery of standard stress tasks or left in a control condition in random order. The Zxcvbn password strength was compared across conditions. Results. We did not observe a statistically significant difference in mean Zxcvbn password strengths on fear (Hedges' $g_{\mathsf{av}} = -0.11$, 95\% CI $[-0.45, 0.23]$) or stress (and control group, Hedges' $g_{\mathsf{av}} = 0.01$, 95\% CI $[-0.31, 0.33]$). However, we found a statistically significant cross-over interaction of stress and TLX mental demand. Conclusions. While having observed negligible main effect size estimates for incidental fear and stress, we offer evidence towards the interaction between stress and cognitive effort that vouches for further investigation.

2020-09-25T12:02:08Z Open Science Framework: https://osf.io/3cd9h. 23 pages. This is the author's technical-report copy. This work was supported by the ERC Starting Grant Confidentiality-Preserving Security Assurance (CASCAde), GA no 716980 Proceedings of the 7th Workshop on Socio-Technical Aspects in Security and Trust (STAST'17), ACM Press, December 2018, pp. 3-15 Tom Fordyce Sam Green Thomas Groß 10.1145/3167996.3168000