INDEX

Explanations

separating opposing parties

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

for

-2.33

to

-2.16

by

-2.05

at

-1.88

on

-1.74

！”

-1.67

 when

-1.67

but

-1.61

！」

-1.60

in

-1.55

POSITIVE LOGITS

ほとん

1.81

 ALREADY

1.73

 refroid

1.69

 protože

1.66

 Trouver

1.66

튿

1.63

々が

1.61

あぁ

1.59

offers

1.59

おそ

1.59

Activations Density 0.012%