INDEX

Explanations

punctuations implying pause

This neuron activates on ellipsis markers (the “…” token).

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

न

1.37

на

1.36

י

1.32

т

1.26

arono

1.09

 troup

1.06

ди

1.03

тли

1.02

н

1.02

स

1.02

POSITIVE LOGITS

?...

1.40

..?

1.36

 ähm

1.27

 sooo

1.26

 seguros

1.23

लेकिन

1.21

 nhưng

1.19

!?

1.18

yeah

1.18

 \\..

1.18

Activations Density 0.054%