Couple of other more businessy reasons:
- SynthID hides the watermark in the sampling RNG. No randomness -> no watermark.
- If you want to distil on the model outputs, you want temp=0 outputs. No temp=0 -> worse distillation.
- SynthID hides the watermark in the sampling RNG. No randomness -> no watermark.
- If you want to distil on the model outputs, you want temp=0 outputs. No temp=0 -> worse distillation.
Certain open models have/had the temperature locked on the official APIs, I assume they just have sampling incompatible with static temperature or do some fancy speculative decoding. It's clearly not to hide anything, as the weights are open and there are always alternative providers.