atomic-level evaluation

llm generation (1) persona fidelity (1) out-of-character behavior (1) large language model (1) persona consistency (1) open-ended generation (1) human-ai interaction (1) evaluation framework (1) text generation (1)