INDEX

Explanations

classifications into groups or types

The neuron activates on words that enumerate or classify items, especially the numerals “two” and “three” when they introduce category counts.

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

-2.42

 aufgrund

-1.52

due

-1.46

ів

-1.43

to

-1.42

nością

-1.42

穽

-1.40

ਞ

-1.39

 hábito

-1.35

 them

-1.34

POSITIVE LOGITS

ต่างๆ

1.50

two

1.36

 elementy

1.34

 kuhusu

1.32



1.30

 сначала

1.29

 oraz

1.27

</u>

1.27

 Appellants

1.26

 bruger

1.26

Activations Density 0.115%