3 ms·
Large Language Models Don’t “Hallucinate”
- PaulHoule 4y agoThat's the point. These models are doing the same thing when they produce accurate output as when they produce inaccurate output, it's just that the strategy they use produces good answers sometimes and bad answers some other times.