Yeah the calibration is really what makes it useful in practice for quick, small decisions. Asking a LLM to give scores to a problem will yield inconsistently scaled/anchored results that changes at a whim.
The blog is pretty heavy on statistics. I'll have to study it more when I have time. Is it essentially bootstrapping results to statistically normalize the answers?
tnspacetime 7 hours ago [-]
I have not studied it properly too. Good that it has both code and note though.
jackb4040 8 hours ago [-]
[flagged]
Rendered at 00:04:22 GMT+0000 (Coordinated Universal Time) with Vercel.
Is this another Claude-ism? "X was always Y. The Z merely hid it." Or am I overcalling it?
[1] https://www.e3displays.com/reflective-lcd-display-monitor/
https://software.human-tokens.dev/
The blog is pretty heavy on statistics. I'll have to study it more when I have time. Is it essentially bootstrapping results to statistically normalize the answers?