INDEX

Explanations

number of

The neuron responds to quantitative expressions, especially “X of Y” constructions where X is a quantity word (e.g. number, amount, lot, area) and Y is a numeric value.

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

 natures

-1.43

隈

-1.38

 quizás

-1.34

 düşük

-1.32

cuerdo

-1.30

 nachdem

-1.30

THOOK

-1.30

ALTH

-1.30

 juxtaposition

-1.28

")){

-1.27

POSITIVE LOGITS

デメリット

1.59

توضی

1.59

帶來

1.55

</h4>

1.53

罟

1.50

 dengan

1.49

مشخص

1.46

حتی

1.45

البته

1.45

 brillo

1.42

Activations Density 0.039%