INDEX

Explanations

"NO." seguido de número

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

</h2>

-2.30

</td>

-1.91

}{

-1.76

-1.71

</h3>

-1.71

<bos>

-1.63

︲

-1.60

 健身

-1.57

，

-1.55

le

-1.52

POSITIVE LOGITS

↵

1.92

the

1.76

 istimewa

1.68

"",

1.62

Ἢ

1.60

鏨

1.58

 kunjung

1.58

it

1.57

in

1.47

 menikmati

1.45

Activations Density 0.001%