INDEX

Explanations

citations and references

The neuron detects numeric citation and reference markers (e.g. volume, page, and other numeric identifiers) in scholarly entries.

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

).

-1.45

IERC

-1.33

Ⅰ

-1.27

。

-1.25

нью

-1.25

of

-1.21

 nothing

-1.20

 multicolore

-1.20

final

-1.19

Login

-1.19

POSITIVE LOGITS

and

2.66

และ

1.41

nsylvania

1.33

ेशा

1.32

 jednod

1.32

 innych

1.28

hoenix

1.27

amd

1.25

补贴

1.23

1.22

Activations Density 0.002%