Benchmark RadarAI BENCHMARK PROFILE
MOSAIC
Adversarial AIGC text detection benchmark with 16,000 samples spanning a full-granularity attack spectrum.
- Released
- 2026-08-28
- Readiness
- Runnable
- Primary field
- General AI
Why it matters
Addresses vulnerabilities of global statistical and semantic-only detectors by providing token-level adversarial scenarios for robust machine-generated text detection.
Motivation
The rapid evolution of large language models necessitates robust machine-generated text detection.
Primary resources
Benchmark Radar records only publicly supported details and links back to primary sources for verification.