INDEX

Explanations

number followed by word/symbol

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

 tasarım

1.59

 primjer

1.58

参数向量中

1.57

 adaptador

1.57

 théâtre

1.56

 テーマ

1.55

∂</

1.55

 główn

1.52

 illustrazione

1.51

 έργ

1.51

POSITIVE LOGITS

<start_of_image>

1.52

↵↵

1.49

↵

1.25

<h2>

1.15

<eos>

1.12

valued

1.10

↵↵↵↵

1.08

brake

1.06

er

1.04

viable

1.02

Activations Density 0.001%