INDEX

Explanations

items and their properties

The main thing this neuron does is detect numeric tokens (i.e. numbers or count values) 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

brtc

-1.53

unno

-1.40

zzet

-1.37

ruka

-1.32

 mukaan

-1.23

ḿ

-1.22

 inversiones

-1.21

celand

-1.20

montage

-1.20

 („

-1.20

POSITIVE LOGITS

 that

3.48

 with

2.50

 which

2.31

can

2.14

may

2.11

of

1.98

 will

1.91

 must

1.88

 cannot

1.84

for

1.73

Activations Density 0.249%