INDEX

Explanations

tabs, code formatting

The neuron detects technical configuration directives (e.g. editor modelines like “tabs” and markup settings like “no index”).

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

-1.45

 which

-1.45

 your

-1.44

 these

-1.30

 Here

-1.28

or

-1.27

 found

-1.23

ものです

-1.20

 this

-1.19

 will

-1.19

POSITIVE LOGITS

<bos>

1.33

instead

1.24

 istället

1.18

 ytterligare

1.16

EVERY

1.13

ovať

1.11

🟢

1.10

äm

1.10

ジナル

1.07

lève

1.07

Activations Density 0.000%