Codú
‹ Back to feed

// Hacker Noon · 12 February 2026

Float vs Int Confidence Scores: Why LLM Output Format Changes Model Behavior

A minimal experiment shows LLM confidence outputs shift with number format: decimals are more conservative and consistent, while some models break in 0–100 mode.

Hacker Noon
@hacker-noon · Felix
hackernoon.com
Read Full Article at hackernoon.com
Hacker Noon@hacker-noon

Discussion 0

Loading

Got something to say?

or to join the conversation.

Learn to build with AI and grow with people doing the same — it's free.