INDEX

Explanations

candidates, history, election results

The neuron strongly activates on numeric tokens—especially standalone numbers and decimal figures—i.e. it detects numerical data in the text.

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

 these

-2.03

of

-1.70

kimono

-1.59

collier

-1.55

 latter

-1.48

 this

-1.45

prestige

-1.45

sterling

-1.43

tapis

-1.43

nikon

-1.42

POSITIVE LOGITS

onomies

1.81

</b>

1.71

antism

1.67

autres

1.62

龶

1.60

1.55

ícia

1.55

cn

1.45

amos

1.44

！”

1.44

Activations Density 0.385%