JEV is just calibrated classification over a label set, not deterministic output
tzmartin · x · 2026-09-23
A skeptical take on the JEV hype: JEV remains probabilistic — a distribution over a fixed set of answers plus a confidence score, not deterministic output. At best it's "calibrated classification" over a user-defined label set, not ranked output in stochastic systems. Cool idea, but the hype is marketing.
More from Models
- GPT-6 Sol accused of ignoring configured skills that work fine on GPT-5.6 — Aber-so-richtig · 2026-09-23
- Opus 5.5 ultra wows with 4-agent, 90-minute fully-from-scratch build — repligate · 2026-09-23
- No Kimi release this week, says leaker: Moonshot staff go on holiday Friday — ChrisGPT · 2026-09-23
- Screenshot suggests GPT-5.6 Sol is free for ChatGPT Go subscribers on Windows — sam619007 · 2026-09-23
- Opus 5.5 wins the day as GPT-6 Sol skips ChatGPT, sparking model debate — koltregaskes · 2026-09-23
- Opus 5.5 becomes a daily driver: faster, cheaper than Opus 5, plus reset credits — addyosmani · 2026-09-23