INDEX

Explanations

database field, sound, acronyms, proper names

The neuron strongly activates on numeric tokens or numeric expressions (i.e. measurements, scale values, counts, etc.).

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 rése

-1.34

considering

-1.34

 magnific

-1.29

hoge

-1.29

 pourraient

-1.28

 trám

-1.28

 SPECIFIC

-1.25

 zoude

-1.24

Pozdrawiam

-1.23

—¡

-1.23

POSITIVE LOGITS

to

1.45

 that

1.34

 first

1.30

 uses

1.28

 used

1.24

as

1.23

 what

1.21

an

1.17

 using

1.16

An

1.15

Activations Density 0.162%