ACL Findings benchmark for measuring LLM sycophancy and correction selectivity
nlp benchmark acl language-models reproducibility ai-safety llm-evaluation sycophancy inspect-ai acl-2026 correction-selectivity
-
Updated
Jul 8, 2026 - Python