INDEX

Explanations

publications/scientific texts

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Publishing

-0.78

ussis

-0.68

SharedCtor

-0.64

ValueStyle

-0.64

Publishing

-0.59

 Bells

-0.59

zebra

-0.59

 noqa

-0.59

iempos

-0.57

weile

-0.56

POSITIVE LOGITS

onAttach

0.55

dom

0.47

 elimination

0.46

ny

0.44

eav

0.43

rager

0.42

oton

0.42

El

0.42

poca

0.42

aving

0.42

Activations Density 0.030%