INDEX

Explanations

string manipulation and data handling

This neuron activates on programming identifiers—class names, method names, and other code tokens (e.g. camel- or Pascal-case API calls and type names).

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

↵↵

-1.79

 contributes

-1.55

࿆

-1.38

↵↵↵

-1.37

</h1>

-1.30

 Tugend

-1.26

 misma

-1.24

-1.23

 discovered

-1.22

ͯ

-1.21

POSITIVE LOGITS

Especially

1.27

Ԁ

1.25

Literally

1.23

様々

1.23

Finally

1.22

损坏

1.22

1.20

 that

1.18

vádz

1.18

 ofrecen

1.17

Activations Density 0.073%