INDEX

Explanations

Questions and code pages

The neuron fires when it sees the identifier “Pages” in code (e.g. as a namespace or folder name).

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

 from

-1.48

who

-1.48

or

-1.41

ؼ

-1.37

for

-1.34

cuba

-1.34

 wybrać

-1.33

になりたい

-1.32

 zrobić

-1.26

marino

-1.24

POSITIVE LOGITS

があります

1.34

 entw

1.30

があり

1.29

Pontos

1.28

Berikut

1.23

Atra

1.21

Secara

1.19

Proces

1.15

 EIGHT

1.15

 intertw

1.09

Activations Density 0.000%