INDEX

Explanations

when needed or necessary

The neuron responds to occurrences of the word “necessary,” especially as part of the phrase “if necessary.”

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

",

-1.40

 gradualmente

-1.33

"},

-1.31

だろ

-1.31

Dank

-1.30

</h1>

-1.30

leggio

-1.25

陆续

-1.23

 Ficha

-1.23

</h5>

-1.23

POSITIVE LOGITS

 каждой

1.47

 sufrimiento

1.38

 chaque

1.37

我們

1.35

 further

1.34

 الأخرى

1.30

pernicus

1.30

 stehe

1.29

out

1.29

ly

1.28

Activations Density 0.013%