INDEX

Explanations

end of list items

The neuron is looking for in-text reference markers—i.e. bracketed citation numbers like “[1]”, “[2]”, etc.

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

 Each

-2.09

By

-1.94

 With

-1.93

 ด้วย

-1.88

 Before

-1.88

If

-1.88

↵↵

-1.74

-1.73

 Several

-1.73

 geomé

-1.69

POSITIVE LOGITS

↵

1.91

鹩

1.80

formatie

1.78

ellido

1.77

 zrobić

1.74

遢

1.74

1.73

 marqués

1.68

 случаях

1.66

癭

1.66

Activations Density 0.019%