INDEX

Explanations

yearlong exercise

species, certain other words

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 Caval

0.50

Ag

0.49

é

0.49

RIC

0.48

TE

0.48

Ic

0.46

for

0.46

Pe

0.46

Am

0.46

 primaire

0.45

POSITIVE LOGITS

mr

0.52

鵡

0.51

キレイ

0.50

噱

0.50

Special

0.50

ParseException

0.49

Certain

0.49

Department

0.47

쇤

0.47

Species

0.47

Activations Density 0.000%