INDEX

Explanations

possessive 's' before divine or academic nouns

The neuron strongly activates on the possessive ’s marker (i.e. words ending in ’s), so it is detecting English genitive ’s forms.

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

ogues

-2.14

itively

-2.13

-1.97

natured

-1.84

byes

-1.79

ergies

-1.77

ologist

-1.73

 preguntó

-1.71

𖡼

-1.69

彣

-1.68

POSITIVE LOGITS

↵

1.91

1.90

and

1.80

dė

1.64

Again

1.59

',

1.59

}^{

1.59

仍然

1.58

1.53

",

1.51

Activations Density 0.009%