INDEX

Explanations

any, anything, anyone

This neuron detects explicit mentions of a person’s gender—particularly statements saying someone “were a female.”

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

และ

-1.35

 некоторые

-1.34

 некоторых

-1.29

 certains

-1.27

 вже

-1.22

and

-1.20

paravant

-1.20

そして

-1.18

 alcuni

-1.16

albeit

-1.16

POSITIVE LOGITS

or

3.81

any

3.48

 anything

2.67

有任何

2.33

 anyone

2.03

 cualquier

1.89

 quelconque

1.87

 или

1.82

 herhangi

1.80

 this

1.78

Activations Density 0.039%