INDEX

Explanations

organization names after 'for'

This neuron activates on numerical tokens (whole numbers or decimals), i.e. it detects numeric values 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

 remnants

-1.60

܆

-1.58

 mixture

-1.58

ఽ

-1.55

icei

-1.55

俛

-1.54

 mystical

-1.53

傒

-1.52

 Dinas

-1.52

耵

-1.52

POSITIVE LOGITS

at

2.38

of

1.94

 from

1.53

 also

1.52

 these

1.50

in

1.43

 found

1.36

for

1.35

who

1.31

1.20

Activations Density 0.068%