INDEX

Explanations

rate, pace, and level

The neuron detects phrases expressing rates, speeds, or other quantitative measures (e.g., “rapid pace,” “high speed,” “acceptable level”).

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

 motocicletas

-1.70

’:

-1.66

过来了

-1.65

从来

-1.59

がありますが

-1.57

繋がり

-1.51

激しい

-1.49

 entregas

-1.49

 strö

-1.48

 fieltro

-1.48

POSITIVE LOGITS

of

2.08

на

1.93

1.86

"";

1.78

it

1.60

."

1.59

en

1.59

''

1.56

н

1.55

""

1.51

Activations Density 0.114%