The case for running the largest model at Max reasoning effort
One commenter sees a largely linear increase in quality as reasoning effort rises.
TLDR
A commenter says they run only the largest model and only at Max effort, citing what they describe as a largely linear increase in quality with reasoning effort.
Combined views
The case for running the largest model at Max reasoning effort
One commenter sees a largely linear increase in quality as reasoning effort rises.
TLDR
A commenter says they run only the largest model and only at Max effort, citing what they describe as a largely linear increase in quality with reasoning effort.
