INDEX

Explanations

multi-million/billion dollar costs

The neuron flags tokens that convey or accompany large quantitative values (e.g. “million,” “billion,” “dollar”) or numerical/statistical contexts (scores, counts, measures).

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

Gambas

-1.73

 möjlighet

-1.72

 promocional

-1.69

还真

-1.65

还真是

-1.59

 interessantes

-1.52

 DÍA

-1.52

憮

-1.51

元々

-1.50

-1.49

POSITIVE LOGITS

 only

1.80

 there

1.61

يتم

1.56

1.52

thargy

1.51

/>\

1.49

"@/

1.45

撼

1.44

 same

1.42

 last

1.41

Activations Density 0.036%