INDEX

Explanations

initially, as with, requires, comparison

like comparisons

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

 Serializable

0.38

Ⅶ

0.38

 STILL

0.38

 HUGH

0.38

OX

0.36

 maxSum

0.36

 still

0.36

పూ

0.36

newcommand

0.36

VII

0.35

POSITIVE LOGITS

まずは

0.47

 एकदा

0.44

最初は

0.43

 требует

0.43

 作り

0.42

 ابتدا

0.42

 रम

0.41

 respectivos

0.40

ifer

0.39

 требуется

0.39

Activations Density 0.297%