INDEX

Explanations

Caucasian Sketches

This neuron primarily detects sentence boundaries, activating on end‐of‐sentence punctuation (commas, periods, ellipses) and the capitalization that marks the start of new sentences.

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

 السياس

-1.64

 agricultural

-1.50



-1.40

$\%$

-1.36

\&

-1.30

\&

-1.29

 ویتامین

-1.22

𝒙

-1.22

in

-1.20

use

-1.20

POSITIVE LOGITS

 Japão

1.47

鉺

1.36

 vanille

1.35

盉

1.31

 quelle

1.27

◖

1.26

 للاسماء

1.25

pherals

1.25

eliness

1.24

諡

1.21

Activations Density 0.027%