INDEX

Explanations

all followed by time

The neuron fires strongly on the quantifier “all,” marking occurrences of “all” used as a universal or exhaustive modifier.

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

ございます

-1.59

 няма

-1.55

is

-1.52

 récemment

-1.45

 désormais

-1.41

 schoenen

-1.38

this

-1.37

 inmigrantes

-1.36

 páns

-1.36

 інші

-1.35

POSITIVE LOGITS

</h5>

1.80

 until

1.69

);

1.60

but

1.53

 years

1.48

 throughout

1.47

The

1.46

</b>

1.46

 year

1.46

 week

1.45

Activations Density 0.018%