04:00
2026-07-07
arxiv.org
large-language-models
Seduced by the Narrative: Assessing Rule Adherence in Semi-Open Textual Sandboxes
Researchers introduced CoC-Seduce, a benchmark for testing LLM rule adherence in semi-open textual environments, revealing that frontier models like GPT-5.4, Claude Sonnet 4.6, and Gemini 3.5 Flash ar…