INDEX

Explanations

to verb

The neuron activates primarily on numeric tokens (quantities or measurements) 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

ところは

-1.69

it

-1.60

 autre

-1.57

by

-1.55

鞴

-1.53

際には

-1.48

拥抱

-1.47

 machete

-1.46

 травня

-1.45

Вита

-1.45

POSITIVE LOGITS

is

1.88

finally

1.63

感觉

1.54

be

1.54

ですが

1.48

न

1.42

夏季

1.41

EdgeInsets

1.40

Were

1.39

Activations Density 0.027%