INDEX

Explanations

names or technical terms

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

 Dana

-0.58

Theorem

-0.58

Dana

-0.53

 Theorem

-0.49

ượu

-0.48

DebuggerStep

-0.48

aticano

-0.47

Irm

-0.46

 definition

-0.46

èbres

-0.46

POSITIVE LOGITS



0.61

Zeneca

0.60

wnicy

0.56

udur

0.52

IsContent

0.50

igshid

0.50

:\/\/

0.48

ValueStyle

0.47

 tetanus

0.47

bibitem

0.45

Activations Density 0.027%