Back to explore

Jev vs GPT-5.4 zorlu fact-check'te: anlamlı fark yok, maliyet 1/50

Topluluk yeniden çalıştırmasında Jev ve GPT-5.4 istatistiksel olarak ayırt edilemiyor (3,3 puan fark) ve Jev'in maliyeti yaklaşık ellide biri.

I said Jev beat GPT-5.4 on hard fact-checks by 11.1 points. That was one run. Rerun five times, I can't tell them apart: 3.3 points, not significant. Jev costs a fiftieth as much. Raw runs: https://github.com/TheWayWithin/jev-bench… https://jamiewatters.work/journey/jev-ties-gpt-5-4…

Jev vs GPT-5.4 zorlu fact-check'te: anlamlı fark yok, maliyet 1/50 1
· 0 likesOpen on X