INDEX

Explanations

code definitions and structures

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

陆续

-1.98

如果

-1.87

 ビキニ

-1.83

 certains

-1.80

渼

-1.76

Hopefully

-1.73

ኧ

-1.73

óricos

-1.71

侂

-1.70

 verschillende

-1.70

POSITIVE LOGITS

↵↵

2.17

in

2.08

for

1.95

 otros

1.82

1.77

 that

1.64

 going

1.63

we

1.62

 before

1.60

 after

1.59

Activations Density 0.014%