INDEX

Explanations

code imports imports and definitions

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

随便

1.06

 ridiculously

1.04

 hopelessly

0.96

enas

0.94

хой

0.88

 horribly

0.85

finement

0.84

 किंवा

0.84

 funny

0.84

 silly

0.83

POSITIVE LOGITS

有助于

1.64

 góp

1.60

有利于

1.48

使得

1.43

 outcomes

1.43

 vital

1.43

 fondamentali

1.42

 crucial

1.41

 enables

1.41

 motivates

1.41

Activations Density 0.494%