INDEX

Explanations

the word "tell"

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 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

 tell

-2.58

 Tell

-2.22

Tell

-2.13

 telling

-2.11

 Telling

-1.98

 tells

-1.96

 TELL

-1.94

tell

-1.92

telling

-1.83

tells

-1.77

POSITIVE LOGITS

 CreateTagHelper

0.64

!*\

0.56

 propOrder

0.53

 OMITBAD

0.52

 comerciais

0.52

 ويكيميديا

0.50

calipsis

0.49

 Arhivirano

0.49

 Réponses

0.48

참고

0.47

Activations Density 0.827%