Yeah, it’s also useful for prompt tuning, debugging and understanding how a model interprets your prompt. Also really good for identifying any contradictions i

Yeah, it’s also useful for prompt tuning, debugging and understanding how a model interprets your prompt. Also really good for identifying any contradictions in your system prompt and context.