INDEX

Explanations

informal colloquial language

The neuron strongly activates on punctuation marks, numeric tokens, and other non-word delimiters (commas, periods, dashes, numbers, etc.).

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 skall

1.42

 loosing

1.39

 नही

1.34

 মোঃ

1.23

 retarded

1.23

 আক্রম

1.20

 כאשר

1.20

 melaksanakan

1.16

出来ない

1.13

 परन्तु

1.11

POSITIVE LOGITS

 staffers

1.29

 voilà

1.16

—

1.16

 unsurprisingly

1.15

 reimag

1.12

 แถม

1.11

 tweaks

1.09

 perks

1.07

 enggak

1.05

Activations Density 2.848%