INDEX

Explanations

code snippets

This neuron appears to be effectively inactive—it does not reliably detect or activate on any pattern.

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

 that

-1.66

بله

-1.59

 теперь

-1.51

fristi

-1.48

now

-1.37

-1.36

„

-1.27

的一次

-1.26

bie

-1.24

なんです

-1.24

POSITIVE LOGITS

ла

1.52

 Therefore

1.51

inverno

1.47

 նաև

1.47

 Thus

1.45

憮

1.45

 TAMBÉM

1.41

and

1.41

"}}

1.40

たと

1.39

Activations Density 0.007%