INDEX

Explanations

do you questions

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 permita

0.88

Punch

0.85

 permanece

0.85

 मेह

0.84

 stitching

0.83

Cra

0.83

 местах

0.82

 paris

0.82

 menyebabkan

0.81

博物

0.80

POSITIVE LOGITS

 certainly

1.47

 definitely

1.32

 DEFIN

1.10

 surely

1.09

俨

1.03

 sure

1.00

显然

0.99

 Certainly

0.99

 defin

0.98

 zeker

0.97

Activations Density 0.107%

No Known Activations