INDEX

Explanations

package names and versions

The neuron strongly activates on file‐ or resource‐name tokens containing extensions (e.g. “.tgz”, “.jpg”) in URLs or paths.

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

for

-1.36

 might

-1.26

or

-1.26

!!!!!

-1.24

 after

-1.23

From

-1.18

With

-1.18

There

-1.17

 just

-1.16

If

-1.16

POSITIVE LOGITS

Gön

1.41



1.34

didn

1.31

綃

1.30

 заявил

1.29

 proie

1.29

besides

1.28

indicate

1.27

featured

1.27

aside

1.25

Activations Density 0.000%