INDEX

Explanations

completing code structures and phrases

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

干嘛

-0.90

って思

-0.89

 conciertos

-0.89

 イベント

-0.88

 prodej

-0.87



-0.85

 나는

-0.84

 démocratie

-0.84

分かります

-0.84

 機能

-0.83

POSITIVE LOGITS

of

1.16

BeforeAll

1.05

 Prise

1.03

 '.';

1.00

 también

0.93

 yaprak

0.91

 ['./

0.89

ptăm

0.89

𝒐

0.88

_{\

0.87

Activations Density 0.007%