INDEX

Explanations

combined with / coupled with

The neuron flags instances of connective phrases where “with” is paired with coordinating words (e.g. “combined with,” “coupled with,” “together with”).

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

’

-1.72

 when

-1.65

 hvid

-1.39

啻

-1.32

if

-1.27

one

-1.27

 costumi

-1.25

甍

-1.25

 térmico

-1.24

 while

-1.23

POSITIVE LOGITS

 that

1.88

栊

1.42

 značky

1.39

媪

1.36

 rumored

1.35

 stolet

1.34

 $^{\

1.34

蓠

1.33

想必

1.30

他の

1.29

Activations Density 0.058%