INDEX

Explanations

abbreviations followed by units

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

푔

-1.73

先週

-1.64

琑

-1.60

 whoſe

-1.59

新しく

-1.59

最近は

-1.58

ziem

-1.57

我们将

-1.56

 motocicleta

-1.56

 semblent

-1.55

POSITIVE LOGITS

or

2.22

if

1.93

This

1.90

for

1.87

These

1.85

 This

1.81

 then

1.81

in

1.80

 with

1.77

 When

1.77

Activations Density 0.014%