4 ms·
That is extremely hard, It require the model to have "knowledge" so it can decide if it know the answer or not which is not how current llm/gpt works
by vdfs 1y ago
That is extremely hard, It require the model to have "knowledge" so it can decide if it know the answer or not which is not how current llm/gpt works