INDEX

Explanations

when something takes effect

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

<i>

-1.53

does

-1.43

 используется

-1.41

\[

-1.37

-1.34

why

-1.34

<blockquote>

-1.34

晤

-1.33

",

-1.31

</strong>

-1.30

POSITIVE LOGITS

for

1.49

 their

1.45

他們的

1.45

۾

1.44

 pubblicato

1.41

on

1.38

مقدمه

1.38

 until

1.36

它们的

1.33

 їх

1.32

Activations Density 0.019%