INDEX

Explanations

all

This neuron responds to words marking universal/general statements—especially the quantifier “all” and its accompanying copula “are” or similar auxiliaries.

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

 also

-2.02

 também

-1.46

 también

-1.46

 نیز

-1.42

 рассказать

-1.31

 també

-1.25

 alfo

-1.25

Emerald

-1.20

 également

-1.17

缆

-1.14

POSITIVE LOGITS

After

1.54

Everybody

1.45

You

1.38

Everyone

1.36

Both

1.33

 ursprünglich

1.32

這

1.29

But

1.27

ここは

1.27

卣

1.26

Activations Density 0.079%