INDEX

Explanations

place, stuff, man

The neuron strongly activates on general placeholder or subject nouns (like “thing,” “place,” “people,” “man”) when they head a descriptive “is a” copula construction.

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

of

-1.55

ꪔ

-1.50

 légèrement

-1.43

婠

-1.34

 innych

-1.33

ᾠ

-1.33

those

-1.30

 betyd

-1.29

땃

-1.27

 mudanças

-1.25

POSITIVE LOGITS

1.70

</strong>

1.57

te

1.32

on

1.31

ować

1.30

 seems

1.30

on

1.27

！」

1.27

 liquor

1.27

 orchestra

1.26

Activations Density 0.024%