INDEX

Explanations

code snippets with parentheses and vertical bars

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

GEBURTSDATUM

-0.96

Hozzáférés

-0.90

verwijspagina

-0.84

 Huntingdon

-0.83

mobileqq

-0.82

﴾

-0.81

 Yorke

-0.80

 Gantt

-0.80

 Kabul

-0.78

Rüyada

-0.75

POSITIVE LOGITS

}|

1.06

)|

0.94

 Hild

0.77

 يتيمه

0.76

ckley

0.75

 Whittaker

0.71

 integri

0.69

ägg

0.69

 Bragg

0.67

rsiniz

0.66

Activations Density 0.006%