INDEX

Explanations

each something, continueorcontinue phrase (e.g., each square, not because)

file and pin descriptions

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

 reproducing

0.48

 Publicity

0.46

Quy

0.45

 publicity

0.45

っぽい

0.45

Nx

0.45

Eb

0.44

MND

0.44

 reproducibility

0.44

 Bers

0.43

POSITIVE LOGITS

いろんな

0.54

전

0.52

号码

0.52

ো

0.49

 것이다

0.49

ወት

0.48

疾

0.47

사

0.46

肩

0.46

िक

0.45

Activations Density 0.001%