INDEX

Explanations

using

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Rhestr

-0.75

Portale

-0.73

 purpoſe

-0.67

#![

-0.67

 protoimpl

-0.65

styleType

-0.63

 مرئيه

-0.63

glGen

-0.62

GOTREF

-0.61

providedIn

-0.61

POSITIVE LOGITS

upsi

0.57

 cliquer

0.52

 Independ

0.48

 surla

0.47

 nhat

0.46

Towards

0.46

gwick

0.46

aspectj

0.46

 Part

0.46

BindView

0.46

Activations Density 0.001%