INDEX

Explanations

words or parts of words that are emphasized

say "s word"

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

Negative Logits

ما

-0.36

์

-0.35

TRIBUN

-0.34

ה

-0.33

elementType

-0.33

SizeF

-0.31

м

-0.30

هم

-0.29

ized

-0.28

getEntity

-0.28

POSITIVE LOGITS

ship

1.57

ski

1.54

san

1.50

sho

1.49

sch

1.49

sing

1.43

shop

1.41

spa

1.41

sen

1.38

sun

1.38

Activations Density 17.918%