INDEX

Explanations

number followed by years

The neuron detects numeric expressions indicating a span of years (a number immediately followed by “years”).

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

by

-2.23

 they

-1.95

 from

-1.82

are

-1.74

 first

-1.57

for

-1.53

all

-1.52

 most

-1.51

 наиболее

-1.48

is

-1.46

POSITIVE LOGITS

*/

1.53

蠓

1.39

 paloma

1.37

 kanker

1.36

 récomp

1.34

 biblioteca

1.34

peut

1.34

"¿

1.33

 dépens

1.33

 europé

1.32

Activations Density 0.028%