INDEX

Explanations

UI elements and code structures

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

 Kindheit

-0.98

 Getränke

-0.96

 recuperação

-0.96

gegevens

-0.93

 Verpflichtung

-0.93

 stylu

-0.90

 Säure

-0.90

 erwor

-0.88

 Verpflich

-0.88

STYLE

-0.87

POSITIVE LOGITS

中小

0.92

ො

0.88

ственные

0.84

蔬

0.84

 substandard

0.83

七年

0.83

抄

0.82

 reinforces

0.81

リア

0.79

玉石

0.79

Activations Density 0.027%