News for hackers. Stuff that matters.

A comment on Microsoft-Decision-1, our model for fast decision-making. Parent comment

Re:Microsoft-Decision-1, our model for fast decision-making by anewhnaccount2 on Friday October 09, 2026 @06:23PM (#50027331)
Yes this is the reason. Transformations like quantisation preserve the rank of outcomes much better than probability mass.
[ Parent | on HN ]
Re:Microsoft-Decision-1, our model for fast decision-making by manmal on Saturday October 10, 2026 @06:02AM (#50031321)
FWIW, my tests used only ranking. I used Kev to drive E2E test runs using agent-device, where Kev had to decide the next action - like pressing a button, or scrolling down etc, to reach a certain goal (UI state). It went surprisingly well, but 4B BF16 was significantly better than 8B Q8. The latter made some wrong decisions in every run while the former was almost at Luna precision.
[ Parent | on HN ]
Re:Microsoft-Decision-1, our model for fast decision-making by _menelaus on Saturday October 10, 2026 @07:07AM (#50031712)
Fascinating insight. I'm curious, how do you know this?
[ Parent | on HN ]