INDEX

Explanations

family members

The neuron activates on words denoting family or kinship relations (e.g., mother, father, husband, wife, daughter).

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

we

-1.80

-1.52

you

-1.52

 więks

-1.41

 natürlich

-1.37

 nossas

-1.31

 they

-1.31

 largely

-1.27

 They

-1.22

of

-1.21

POSITIVE LOGITS



1.46

⎤

1.33

 agua

1.29

 допомогти

1.28

}></

1.27

哟

1.27

enerys

1.22

 mengajar

1.21

}}=

1.20

ступить

1.19

Activations Density 0.036%