INDEX

Explanations

opened up, elicit a

identifying specific trigger words

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 дер

0.48

 EXHIB

0.48

ாலும்

0.47

 Toutes

0.45

 Виктор

0.45

óż

0.44

Ἐ

0.44

 পদ্ধতিতে

0.44

離婚

0.44

ખ

0.43

POSITIVE LOGITS

 मजाक

0.47

eda

0.47

ayah

0.46

qm

0.46

feer

0.45

ida

0.44

 habl

0.44

 tedy

0.44

eco

0.44

इसलिए

0.44

Activations Density 0.003%