INDEX

Explanations

words associated with him/her/them

The neuron activates on tokens naming or referring to people—proper names and personal pronouns.

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

 zelfs

-1.30

かなかった

-1.30

 этого

-1.30

鼙

-1.16

 къде

-1.14

 öppna

-1.13

 gemakkelijk

-1.13

 voldo

-1.12

RESULTADOS

-1.12

 agradec

-1.09

POSITIVE LOGITS

 they

1.84

 mereka

1.60

They

1.53

让他们

1.46

 ihnen

1.44

 their

1.43

 They

1.41

as

1.30

1.25

给他们

1.24

Activations Density 0.158%