INDEX

Explanations

nice vintage, you can

varying execution, often inefficient

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

يلا

0.51

 भारी

0.49

 homebuyers

0.44

Shower

0.43

িরের

0.43

ԁ

0.43

শিক

0.43

长大

0.42

 adelante

0.42

澡

0.41

POSITIVE LOGITS

0.57

 Eduard

0.54

Ep

0.52

All

0.52

 Test

0.51

 Jürgen

0.50

 Jiangsu

0.50

 Prés

0.49

Jir

0.47

Jun

0.47

Activations Density 0.000%