INDEX

Explanations

closing braces and punctuation

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

</h2>

-2.67

-2.44

-2.03

 ytterligare

-1.79

-1.77

见状

-1.74

 vasco

-1.73

 грудня

-1.66

Name

-1.66

’

-1.59

POSITIVE LOGITS

be

2.17

 they

2.05

Of

1.97

Now

1.85

the

1.80

">//

1.73

Questo

1.72

 ciertas

1.72

 numerosos

1.71

 Anſ

1.70

Activations Density 0.034%