It's quirks are too smooth! Very strange. I'm wondering if the effect is due ML models in general (and LLMs in specific) being unable to step outside the bounds

It's quirks are too smooth! Very strange. I'm wondering if the effect is due ML models in general (and LLMs in specific) being unable to step outside the bounds of their training data.