Topic: audit

5 stories found

Wednesday, August 26, 2026

research35

Auditing the Synthetic Memoir: Measuring Scene-Level Confabulation in LLM-Generated Autobiography Against the Documented Record of the Life It Describes

A study audited an AI-generated memoir against factual records, finding instances of confabulation at the scene level. This matters as it highlights the accuracy limitations of large language models in producing autobiographical content.

arxiv.orgโ†—

Monday, August 24, 2026

research40

Exploratory As-Analyzed No-Detection of Culturally-Marked Predicate-Triggered PII Amplification in a Synthetic-English RAG Probe: A Predicate-Resource-Confounded Audit

The study investigates if stereotype-loaded queries reveal more personally identifiable information from a RAG system compared to neutral ones, focusing on culturally marked topics across four cultures. This matters as it highlights potential biases and privacy risks in AI systems when handling sensitive cultural queries.

arxiv.orgโ†—

๐ŸŒฟ That's all for now. Come back tomorrow.

5 of 5 items shown. Sources: 107 days indexed.