INDEX

Explanations

checking for negative values

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Visual

1.24

광역시

1.23

キャッシュ

1.20

Movie

1.17

Timeline

1.17

 unimportant

1.15

 zadr

1.14

 আসেনি

1.13

 যায়নি

1.13

 Spinner

1.12

POSITIVE LOGITS

 Ле

1.82

ູ້

1.80

嚯

1.80

 उदार

1.66

 Кор

1.65

 البريط

1.64

的

1.63

 উল্ল

1.62

ួយ

1.61

 مؤس

1.61

Activations Density 0.010%