INDEX

Explanations

should decide

Numpy, body, shrinking

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

high

0.49

the

0.47

 shooters

0.46

 стрельца

0.44

"{

0.43

djang

0.43

llam

0.43

လ

0.43

zhi

0.43

ٹرز

0.42

POSITIVE LOGITS

 annih

0.48

拡大

0.47

 निवार

0.46

 semelhantes

0.45

 موض

0.45

 aktive

0.45

rer

0.45

 andre

0.45

 capucha

0.45

ammad

0.45

Activations Density 0.001%