Evaluating LLMs with Multiple Problems at once: A New Paradigm for Probing LLM Capabilities

Open in new window