INDEX

Explanations

duga

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

 utafitiHapana

-0.55

mption

-0.53

 hâte

-0.51

IContainer

-0.51

ArgsConstructor

-0.48

ButterKnife

-0.48

 réserver

-0.48

Eksteraj

-0.48

rillation

-0.48

 Extragalactic

-0.48

POSITIVE LOGITS

ValueStyle

0.60

Gid

0.59

>{@

0.59

AnchorTagHelper

0.57

ftagPool

0.56

Карьера

0.54

letoe

0.53

__).

0.53

%)$

0.52

 genetically

0.52

Activations Density 0.065%