INDEX

Explanations

greetings and closings

The neuron detects message salutations that pair a greeting with the word “from” (e.g. “Greetings from,” “Hello from,” “Good Morning from”).

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

-1.54

 such

-1.41

-1.39

you

-1.33

if

-1.32

 which

-1.30

-1.28

or

-1.27

</b>

-1.27

</

-1.26

POSITIVE LOGITS

ými

1.41

sistors

1.36

 ﬁrst

1.35

ПРИ

1.31

可爱

1.31

 ПРЕ

1.30

ージャー

1.29

suma

1.29

spania

1.26

 extrême

1.25

Activations Density 0.008%