INDEX

Explanations

quantifying phrases

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

-0.10

Ł

-0.09

clock

-0.09

linger

-0.09

/////////////////////////////////////////////////////////////////////////////↵

-0.09

the

-0.08

inez

-0.08

 Sabb

-0.08

 dazz

-0.08

dek

-0.08

POSITIVE LOGITS

 sorts

0.15

olec

0.12

 roughly

0.11

 least

0.10

 approximately

0.10

 zero

0.10

anych

0.09

fare

0.09

ungan

0.09

sted

0.09

Activations Density 0.231%