INDEX

Explanations

references to baseball

from

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

 للاسماء

-0.63

APOLIS

-0.58

Autoritní

-0.58

 Lumpur

-0.54

 hypot

-0.52

 barge

-0.49

 tanong

-0.49

 myſelf

-0.49

-0.48

 bại

-0.48

POSITIVE LOGITS

DebuggerStep

0.57

</thead>

0.54

ьаж

0.52

RefNanny

0.50

SourceChecksum

0.48

效

0.48

LookAnd

0.47

gic

0.47

lorum

0.47

Pan

0.46

Activations Density 0.928%