INDEX

Explanations

film credits or resume items

The neuron activates on words and phrases that introduce or quantify a person’s achievements or experience (e.g. counts of credits, years, “more than,” “under,” “belt,” “resume”).

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

 would

-2.05

if

-1.82

 because

-1.77

 while

-1.66

for

-1.62

-1.58

 when

-1.55

 could

-1.52

 like

-1.52

or

-1.50

POSITIVE LOGITS

曖昧さ

1.63

ญี่ป

1.57

榊

1.54

榎

1.52

 convite

1.51

eer

1.51

 silêncio

1.51

 pulseira

1.50

 récolte

1.48

 sparkles

1.48

Activations Density 0.024%