INDEX

Explanations

The neuron predominantly activates on mentions of specific people’s names or personal referents (e.g., pronouns and proper names, especially of women).

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 himself

1.48

 manhood

1.35

 അദ്ദേഹം

1.23

 brotherhood

1.23

 prostatic

1.23

 ছাত্রদের

1.22

 boyhood

1.20

 shirtless

1.19

 manly

1.11

彼は

1.10

POSITIVE LOGITS

 actresses

2.66

 feminina

2.55

 herself

2.48

 женщина

2.42

 businesswoman

2.39

 heroine

2.39

 woman

2.38

 Woman

2.38

 femenina

2.36

 актриса

2.34

Activations Density 2.948%