INDEX

Explanations

past actions and results

The neuron fires on past-participle or passive-voice verb forms (e.g. “made,” “expressed,” “shown”).

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

کفش

-1.38

StoryboardSegue

-1.38

ificantly

-1.38

 sagde

-1.31

縈

-1.29

tembro

-1.25

 válvula

-1.22

آیا

-1.21

 них

-1.20

izados

-1.20

POSITIVE LOGITS

of

2.59

to

1.80

1.55

by

1.52

1.40

in

1.29

1.24

by

1.19

的家

1.16

Activations Density 0.058%