INDEX

Explanations

time off or absence

This neuron detects mentions of taking time off (e.g. “off” in the context of days or time away from work).

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

-1.65

—

-1.60

-1.52

 bière

-1.50

 cadeira

-1.45

-1.42

 portefeuille

-1.40

 passar

-1.40

 Think

-1.38

欺负

-1.37

POSITIVE LOGITS

to

2.92

 from

2.34

 being

1.82

ที่จะ

1.80

or

1.63

but

1.61

 ameryka

1.55

 奶油

1.52

 such

1.52

herited

1.50

Activations Density 0.010%