INDEX

Explanations

phrases related to taking action

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

HomeAsUpEnabled

-0.63

subpackage

-0.57

 fragment

-0.52

 wester

-0.51

TBA

-0.50

價值

-0.50

 Esch

-0.50

uchtung

-0.50

Xna

-0.50

Essentially

-0.49

POSITIVE LOGITS

 precautions

0.92

 MEASURES

0.91

 Precautions

0.91

 precaution

0.88

Measures

0.86

 measures

0.85

 Measures

0.85

 precautionary

0.82

 mesures

0.81

measures

0.80

Activations Density 0.037%