INDEX

Explanations

referring to content above

The neuron activates on the word “above” when referring to embedded media or content earlier in the document.

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

 with

-2.16

for

-1.84

is

-1.82

-1.71

by

-1.67

to

-1.67

 from

-1.67

for

-1.59

FOR

-1.52

"};

-1.52

POSITIVE LOGITS

 cosmé

1.97

怍

1.80

 étanche

1.79

𓆏

1.77

 cerro

1.75

 banyo

1.73

 guapos

1.72

 fabrique

1.71

 above

1.70



1.70

Activations Density 0.047%