INDEX

Explanations

please and preferably

instructions or requirements

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

 unwittingly

0.47

 pourtant

0.46

 accidens

0.46

 нередко

0.45

 ternyata

0.44

 unconsciously

0.44

 weakened

0.43

 inadvertently

0.42

 unexpectedly

0.41

变成了

0.41

POSITIVE LOGITS

 preferably

1.14

 Preferably

1.01

 желательно

0.99

preferably

0.99

plz

0.75

 কমপক্ষে

0.75

 please

0.74

 PLEASE

0.73

 helst

0.71

pls

0.70

Activations Density 0.093%