INDEX

Explanations

Technical Language

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

 ujednoznacz

-0.66

terase

-0.62

elemField

-0.59

invokeLater

-0.58

exitRule

-0.58

stateMutability

-0.58

 resourceCulture

-0.57

 onData

-0.57

 dazu

-0.57

cách

-0.57

POSITIVE LOGITS

knapp

0.51

Hai

0.45

 pauper

0.44

 CURIAM

0.43

 dAtA

0.43

pj

0.42

 comm

0.42

 commode

0.41

MAR

0.41

ęło

0.41

Activations Density 0.004%