INDEX

Explanations

the word "magnitude" and nearby words

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

ErrorListener

-0.74

every

-0.65

Zan

-0.64

“……

-0.64

 Bress

-0.63

Zan

-0.62

 Cress

-0.62

κίνη

-0.62

aronder

-0.60

PreAuthorize

-0.60

POSITIVE LOGITS

Magnitude

1.31

 magnitude

1.30

 Magnitude

1.30

 magnitudes

1.28

 magnitud

1.13

magnitude

1.12

 Magellan

0.97

MAG

0.93

 mags

0.91

 Maguire

0.88

Activations Density 0.004%