INDEX

Explanations

</strong>

This neuron detects emphasis or formatting markers (e.g. HTML tags like <strong> or <b> and markdown ** bold markers).

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

 célé

-1.39

 zahlreiche

-1.34

ExceptionHandler

-1.34

 возрасте

-1.32

ҍ

-1.29

δος

-1.28

essere

-1.27

 betyder

-1.26

…]

-1.25

 çekilen

-1.25

POSITIVE LOGITS

 Paramètres

1.38

<em>

1.33

上的

1.31

1.30

 shows

1.27

 provided

1.27

 UTIL

1.22

oléon

1.20

1.19

Activations Density 0.008%