INDEX

Explanations

non-latin alphabets

actions across languages

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

0.36

of

0.36

 prudent

0.35

 bone

0.34

 housekeeping

0.34

 tabular

0.34

 stewardship

0.34

 Stewardship

0.34

 regions

0.33

 sediments

0.33

POSITIVE LOGITS

ר

0.50

ד

0.47

ЛИ

0.46

רו

0.46

לי

0.46

אי

0.45

리

0.44

КА

0.43

Те

0.43

ג

0.43

Activations Density 0.150%