INDEX

Explanations

sentence boundary markers

The neuron responds to numeric tokens—years, dates, statistics, percentages, and other numbers—in the text.

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

 hefty

-1.39

 общего

-1.34

 disposto

-1.30

もあった

-1.29

 talvez

-1.28

bkz

-1.27

臯

-1.26

桶

-1.24

 手工

-1.23

 momentous

-1.22

POSITIVE LOGITS

and

2.59

but

1.75

 EVERY

1.61

 because

1.58

as

1.49

 setiap

1.42

 every

1.26

宴

1.21

我们

1.20

 unusually

1.20

Activations Density 0.245%