benchgap
Calibration

Decision Index Language → HLE-MultiChoice (Decision Index)

HLE-MultiChoice (Decision Index) is estimated from Decision Index Language with a Michaelis–Menten curve fitted on 11 models measured on both: y = 2.0000·x / (7.23891 + x), R² = 0.58, cross-validated error 2.4 pp. It is used for 0 estimates.

Fitted on

ModelDecision Index LanguageHLE-MultiChoice (Decision Index)
Clef47.8%11.1%
Clef-Flash40.5%10.3%
Clef-Omni40.7%9.0%
Drex Max62.1%16.0%
Drex v1.540.8%12.1%
Drex v2.052.7%13.6%
elephant-162.1%18.3%
GLiDE v1.266.8%17.5%
Jev 1.1458.0%18.9%
Perplexity Decider Flash v1.2 4B62.2%12.1%
Perplexity Decider v1.2 27B65.5%14.6%