INDEX

Explanations

words related to suggestions and instructions

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

The

-0.57

 itſelf

-0.54

 sportivo

-0.53

 což

-0.52

 Ilustra

-0.51

 Utilizamos

-0.50

 Socialista

-0.50

 realizes

-0.50

 grows

-0.49

 vastaan

-0.48

POSITIVE LOGITS

 Visit

0.85

 yourself

0.80

Visit

0.80

bitField

0.76

Get

0.74

 Check

0.74

Try

0.74

 Explore

0.73

ंदीखरीदारी

0.73

RegressionTest

0.72

Activations Density 1.126%