INDEX

Explanations

descriptive adjectives and verbs

no lasting impact

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

OSE

0.53

钎

0.52

取付

0.51

OW

0.49

 ಬೆಳ

0.49

OUGH

0.48

鞏

0.47

STRU

0.47

subsection

0.46

OCKET

0.46

POSITIVE LOGITS

 smells

0.50

 পারেন

0.46

')).

0.46

 neko

0.45

 کی۔

0.44

 mogę

0.44

 murió

0.41

 biota

0.41

Sibling

0.41

 dazu

0.41

Activations Density 0.000%