Benchmark RadarAI BENCHMARK PROFILE
BenSyc
BenSyc is a benchmark for studying conversational sycophancy in Bengali social contexts, with binary labels and a five-level taxonomy from invalidation to escalation.
- Released
- 2026-06-08
- Readiness
- Paper only
- Primary field
- General AI
Why it matters
It fills a gap in sycophancy research by focusing on culturally grounded conversational alignment in Bengali, important for socially aligned conversational AI.
Motivation
Large language models (LLMs) increasingly participate in emotionally sensitive social conversations, where responses may shift from balanced support toward excessive validation or escalatory alignment.
Primary resources
Benchmark Radar records only publicly supported details and links back to primary sources for verification.