INDEX

Explanations

comment, in, not, now, as, check

alternatives or,

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

 বলছি

0.43

pru

0.40

 impug

0.39

 آینده

0.39

፦

0.39

 welkom

0.38

んにちは

0.37

 каком

0.37

 paheli

0.37

 embe

0.36

POSITIVE LOGITS

which

0.70

Above

0.66

অর্থাৎ

0.66

And

0.65

and

0.63

etc

0.60

Note

0.59

There

0.58

Clearly

0.57

といった

0.57

Activations Density 0.045%