INDEX

Explanations

contrasting with many

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

wie

3.15

𝑙

3.10

ndan

3.02

 männ

2.96

haa

2.90

ismen

2.87

疵

2.83

эле

2.81

ido

2.76

idded

2.71

POSITIVE LOGITS

gomery

3.97

ーション

3.14

2.92

ের

2.86

serrat

2.80

2.79

える

2.69

essori

2.65

enance

2.65

 Pros

2.63

Activations Density 0.009%