INDEX

Explanations

Advancements/Improvements

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

IsContent

-0.74

complexContent

-0.73

seamnă

-0.64

 שוליים

-0.63

 otomatig

-0.61

ագրություններ

-0.60

LEncoder

-0.58

extAlignment

-0.58

 réguli

-0.57

tront

-0.57

POSITIVE LOGITS

the

1.09

our

0.65

 betweenstory

0.50

 intrinsic

0.45

 interviewer

0.43

mouseleave

0.43

+#+#

0.42

 their

0.41

 bonding

0.40

 correlations

0.40

Activations Density 0.001%