INDEX

Explanations

that followed by variety

The neuron activates on common nouns that name abstract concepts or categories (e.g. “challenges,” “paths,” “technologies,” “problems,” “role”).

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

They

-1.70

-1.64

and

-1.52

-1.51

importation

-1.48

-1.38

allure

-1.37

estars

-1.36

เครื่อง

-1.34

стей

-1.31

POSITIVE LOGITS

you

3.42

 they

1.85

our

1.84

we

1.81

of

1.80

 these

1.77

 pulau

1.72

はな

1.70

 your

1.69

袿

1.68

Activations Density 0.085%