Measurement, Detection
No mentions found
This entity hasn't been tracked yet, or Iris is still building its knowledge base.
Related Articles from SNS
Description-Code Inconsistency in Real-world MCP Servers: Measurement, Detection, and Security Implications
arXiv:2606.04769v1 Announce Type: new Abstract: The Model Context Protocol (MCP) has emerged as a critical standard empowering Large Language Models (LLMs) to utilize external tools. In this ecosystem, LLMs rely on natural language descriptions provided by MCP servers to select and execute functions. This interaction implicitly assumes that tool descriptions faithfully reflect their underlying implementations, while this assumption is not mandatorily verified in practice.
Towards AI epidemiology: a measurement standardisation framework for prospective risk detection
Announce Type: replace Abstract: This paper proposes a measurement standardisation framework that compresses expert-AI interactions into structured, comparable fields for prospective risk detection in deployed AI systems, without access to model internals. The main aim of this concept paper is to define the scope of the framework, both semantically and statistically, and to specify a protocol for its empirical testing in future work. The population-level claims the framework is designed to...
Gap-K%: Measuring Top-1 Prediction Gap for Detecting Pretraining Data
Announce Type: replace Abstract: The opacity of massive pretraining corpora in Large Language Models (LLMs) raises significant privacy and copyright concerns, making pretraining data detection a critical challenge. Existing state-of-the-art methods typically rely on token likelihoods, yet they often overlook the gap between the target token and the model's top-1 prediction, as well as local correlations between adjacent tokens. In this work, we propose Gap-K%, a novel pretraining data...
Measurement of the X-ARAPUCA's Absolute Photon Detection Efficiency for the Deep Underground Neutrino Experiment's Vertical Drift Far Detector
arXiv:2511.12328v2 Announce Type: replace Abstract: The DUNE experiment will implement a photon detection system composed of X-ARAPUCA (XA) devices. These trap incoming VUV photons by internal reflection in a wavelength shifter light guide to be collected onto silicon photomultiplier arrays, sensitive to visible light. In the baseline design, dichroic filters are used to prevent photons from escaping.
Optically detected nuclear magnetic resonance of carbon-13 in bulk diamond
Announce Type: replace-cross Abstract: Precision measurements based on optically detected nuclear magnetic resonance offer exquisite sensitivity to absolute shifts in spin transition frequencies, with potential applications in fundamental physics experiments and inertial sensing. We investigate 13C nuclear spins in diamond as a candidate system for solid-state implementations, which hold the promise for high-fidelity readout of large numbers of coherent nuclear spins in millitesla or lower...
Enhancing Hallucination Detection through Noise Injection
arXiv:2502.03799v4 Announce Type: replace Abstract: Large Language Models (LLMs) are prone to generating plausible yet incorrect responses, known as hallucinations. Effectively detecting hallucinations is therefore crucial for the safe deployment of LLMs. Recent research has linked hallucinations to model uncertainty, suggesting that hallucinations can be detected by measuring dispersion over answer distributions obtained from multiple samples drawn from a model.
Thai watchdog to sue Meta over Facebook scam ads targeting users
Thai watchdog to sue Meta over Facebook scam ads targeting users BANGKOK: Thailand's consumer watchdog said it will sue Meta's Facebook for allegedly allowing scammers to use the platform to defraud users through adverts and for failing to protect consumers, the Consumer Council said on Thursday (Jun 4). Meta did not immediately respond to a Reuters request for comment. In previous cases, the company has said it invests in measures to detect and remove scam content and works with regulators...
Topology as Logic: Structural Role Geometry Across Formal, Software, Biological, and Prebiotic Systems
Announce Type: new Abstract: We ask whether dependency topology correlates with functional load-bearing organization as recoverable geometry -- not as a metaphor, but as a measurable structural property detectable by multilayer network analysis. Across seven independent substrates, we show that hub persistence and rank divergence under the Functional Proximity Law recover operational organization that domain experts describe as logic: axiomatic load-bearing structure in formal mathematics,...
Geometry-Driven Flow Analysis of Brain Sulcal Pattern
Announce Type: new Abstract: Cortical folding reflects coordinated neurodevelopmental processes and is increasingly recognized as a sensitive marker of neurological disease. However, most existing analyses rely on indirect scalar summaries that do not explicitly model folding geometry itself. In juvenile myoclonic epilepsy (JME), a common genetic epilepsy, cortical abnormalities are often subtle, spatially distributed, and difficult to detect using conventional morphometric measures.
MalSkillBench: A Runtime-Verified Benchmark of Malicious Agent Skills
Announce Type: new Abstract: AI coding agents such as Claude Code and Gemini CLI increasingly extend themselves with third-party skills: markdown packages bundling natural-language instructions, executable scripts, and tool permissions. Because a skill is at once code and agent-facing instruction, it introduces a supply chain dependency whose risk is neither pure code nor pure prompt. Detection tools have never been measured against verified ground truth spanning this hybrid space, leaving...