arXivMatteo Greco, Anudeex Shetty, Andrea Tagarelli, Jey Han LauWed, Sep 2, 2026, 2:52 AM PDT
score 16.2
New benchmark tests AI attribution across languages and genres
Original: MultiGhostBench: A Multilingual Benchmark for Long-Form LLM-Generated Text Attribution under Distribution Shifts
Source: arxiv.org ↗
Writing ELI5 summary…