arXiv:2602. 18518v2 Announce Type: replace Abstract: Content safety teams need metrics that reflect what users actually experience, not only what is reported.
arXiv:2602. 16111v2 Announce Type: replace-cross Abstract: Online media platforms track the share of impressions associated with content attributes, or prevalence, to evaluate trade-offs and set guardrails in A/B experiments.
arXiv:2605. 17034v2 Announce Type: replace-cross Abstract: Standard PII filters often miss contextual data leakage in RAG systems, such as non-regulated attribute clusters that collectively identify individuals.
The study audits Bluesky’s Moderation Service (BMS) using its 10.6 million public moderation labels from 2025. It finds that BMS operates as a human‑AI collaboration: sexual and graphic content is flagged automatically in seconds, while more nuanced or high‑stakes content requires human review that can take hours or days. The system shows high precision (0.837) but low recall (0.222), with annotators detecting 4.5 times more harmful content than the system, and clustering reveals harms ranging from hostility toward protected groups to the spread of explicit material.
The paper introduces WSF-ARG+, a new dataset that pairs hate speech with check‑worthiness annotations, and presents an LLM‑in‑the‑loop framework to streamline the annotation process. Experiments with 12 open‑weight large language models demonstrate that the framework cuts human effort while maintaining annotation quality. The study also shows that incorporating check‑worthiness labels improves hate‑speech detection performance, boosting macro‑F1 scores for large models by up to 0.213 and averaging 0.154 across models.
EvoHarmBench is a dynamic adversarial evaluation framework that simulates how users iteratively modify harmful content to evade moderation. It uses an optimization loop that evolves evasion strategies at the semantic-cluster level while maintaining human readability, and tests 229 semantic sub-clusters across five violation categories derived from 5,002 real-world adversarial samples. The study shows that even state‑of‑the‑art LLM‑based moderators can be bypassed with an 80.3% success rate after twelve iterations, highlighting significant vulnerabilities in current systems.
arXiv:2604.21549v2 Announce Type: replace Abstract: Estimating the prevalence of a category in a population using imperfect measurement devices (diagnostic tests, classifiers, or large language model...
The paper introduces a training‑free approach to detect policy violations in large language models by treating the task as an out‑of‑distribution problem in the model’s activation space. It uses whitening‑inspired techniques to compute policy‑violation scores directly from normalized hidden activations, requiring only the policy text and a few illustrative examples. Experiments on several LLMs and policy benchmarks show the method achieves up to 86.0% F1, outperforming fine‑tuned and LLM‑as‑a‑judge baselines while being computationally lightweight.
Calpric is a system that combines automatic text selection, segmentation, active learning, and crowdsourced annotation to create a large, balanced training set for privacy policy classification. By simplifying the labeling task, it enables untrained crowd workers to match the performance of trained annotators and reduces inter‑annotator disagreement, cutting labeling costs. The approach yields a dataset of 16,000 policy text segments across nine data categories and produces models that deliver accurate, fine‑grained labels at a cost of roughly $0.92–$1.71 per segment.
arXiv:2401.08038v2 Announce Type: replace-cross Abstract: A significant challenge to training accurate deep learning models on privacy policies is the cost and difficulty of obtaining a large and com...
arXiv:2606. 15029v1 Announce Type: new Abstract: LLM judges are used to reduce the need for costly human labor in evaluating open-ended text generation.
arXiv:2605. 13801v2 Announce Type: replace-cross Abstract: As generative AI models such as large language models (LLMs) become more pervasive, ensuring the safety, robustness, and overall trustworthiness of these systems is paramount.
We study imbalanced crowdsourcing with a focus on class-dependent annotator accuracy, a setting that, to the best of our knowledge, remains relatively underexplored despite its importance in real-world inspection systems where the labels of greatest operational importance are also the rarest ones. In this setting, annotators may be reliable on both classes, unreliable on both classes, majority-class specialists, or minority-class specialists.