INDEX

Explanations

nice

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

-0.48

Wrap

-0.47

 self

-0.46

مصادر

-0.46

longValue

-0.44

-0.43

now

-0.43

 plot

-0.42

 Wrap

-0.42

ύτε

-0.42

POSITIVE LOGITS

matchCondition

0.91

WillAppear

0.91

principalColumn

0.90

DoubleQuotes

0.89

 виправивши

0.86

 gynhyrchwyd

0.86

InjectAttribute

0.84

featureID

0.75

 Himo

0.75

 Roskov

0.75

Activations Density 0.129%