If the model is instructed to periodically ask the user to start from a clean slate context, and some users do comply with that, they probably have good stats on average size of context cache use for users who are presented with that answer (vs users who are not), basic A/B testing stuff.

Might also be performance related in tok/s for what users will perceive as a more speedy experience. For a much smaller scale example, compare local performance of qwen 3.6 27B (not MoE) Q8 with 250,000+ context available, run on local hardware, tok/s generation rate when context is empty vs when context used is at 95,000. Same principle will apply to a much larger model.

Also, it can be riled up to do the task better, or try harder, if I include how cool it will be to get this task done.

But also I feel I have this Orwellian task of censoring its text to avoid it spiralling into negative territory where it convinces itself that the tasks are too difficult. Strange times!