INDEX

Explanations

div, View, Scrollable, Transcript, Margaret

This neuron strongly fires on tokens that name UI or markup elements (e.g. HTML tags or component/view identifiers).

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

and

-1.49

 find

-1.32

 From

-1.25

 this

-1.24

for

-1.20

 These

-1.19

).

-1.16

づけ

-1.15

-1.13

POSITIVE LOGITS

 orquesta

1.40

<bos>

1.35

 asesinado

1.34

Αν

1.31

Bjs

1.28

 aler

1.28

identifikasi

1.25

🩻

1.24

 elektri

1.23

 uova

1.22

Activations Density 0.000%