INDEX

Explanations

dates and numbers

The neuron is strongly activated by four-digit numbers that represent years in date headers.

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

נטרנט

-1.49

aimerais

-1.41

ówczas

-1.39

DNEY

-1.35

Linki

-1.34

 Accesorios

-1.32

 cứ

-1.31

 сыром

-1.26

venser

-1.24

"\\

-1.23

POSITIVE LOGITS

↵

1.66

地说道

1.59

亿元

1.52

 pomocou

1.48

</strong>

1.45

1.38

地说

1.37

mē

1.34

<i>

1.34

 когда

1.30

Activations Density 0.172%