INDEX

Explanations

measurements

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

UserScript

-0.57

<bos>

-0.52

 ویکی‌پدیای

-0.52

Cursors

-0.51

InitVars

-0.50

 referenties

-0.49

Credentials

-0.49

TestTools

-0.47

rót

-0.47

коменду

-0.47

POSITIVE LOGITS

 Chwiliwch

0.56

ChildScrollView

0.55

 rospy

0.53

 تانيه

0.53

hnia

0.52

StructEnd

0.52

RTGC

0.51

MessageTagHelper

0.50

 ExecuteAsync

0.48

brainly

0.48

Activations Density 0.003%

No Known Activations