INDEX

Explanations

in such a

The neuron activates on standalone numeric values (e.g. measurements, section or figure numbers) in the text.

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

providedIn

-1.63

巷

-1.30

OCTOBER

-1.30

nischen

-1.27

 Москву

-1.25

Ответить

-1.25

deutschen

-1.24

}));

-1.23

utdown

-1.22

 complicado

-1.22

POSITIVE LOGITS

 these

1.88

the

1.70

↵

1.69

1.59

 each

1.54

it

1.46

 that

1.41

 Then

1.38

are

1.35

 they

1.32

Activations Density 0.188%