INDEX

Explanations

coming, came, comes

The neuron activates on tokens containing the character sequence “com.”

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

-1.80

 perpé

-1.56

 have

-1.52

 sés

-1.46

 soulign

-1.42

 grecque

-1.38

 véhic

-1.38

 सकते

-1.37

 stolet

-1.37

 dépens

-1.36

POSITIVE LOGITS

ांकि

1.46

</u>

1.38

օ

1.30

」、「

1.29



1.24

 apareció

1.23

▼

1.23

 setuju

1.23

予約

1.19

$=$

1.19

Activations Density 0.022%