INDEX

Explanations

Thursday Oct

understanding concepts, identifying patterns

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

 детства

0.51

னோ

0.48

 casque

0.48

 Ernest

0.47

 Gerardo

0.47

 segundo

0.46

噺

0.46

 admire

0.45

 primeiro

0.45

 Jésus

0.45

POSITIVE LOGITS

0.53

nt

0.51

vs

0.50

사

0.49

in

0.47

naj

0.46

Formal

0.46

 Requirements

0.45

ln

0.45

Activations Density 0.000%