INDEX

Explanations

in

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

webElementXpaths

-0.87

SourceChecksum

-0.85

 للاسماء

-0.80

ValueStyle

-0.79

SharedDtor

-0.75

 ddelweddau

-0.74

 Wikimedijinoj

-0.71

UserScript

-0.71

IsContent

-0.71

 NgModule

-0.71

POSITIVE LOGITS

 general

0.79

so

0.61

general

0.56

 terms

0.54

all

0.48

 combined

0.47

 General

0.46

 unspecified

0.45

 overall

0.44

 essentially

0.43

Activations Density 0.001%