INDEX

Explanations

possessive pronouns followed by concrete nouns

The neuron fires on personal possessive pronouns, especially “my” (and to a lesser extent “your”).

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

 kilku

-1.63

 '*':

-1.59

-1.58

 perioden

-1.55

-1.53

 můžete

-1.49

ムラ

-1.48

 kuning

-1.48

 parle

-1.48

 март

-1.47

POSITIVE LOGITS

by

1.87

 Each

1.86

for

1.76

 Usually

1.68

 Since

1.67

 Almost

1.65

 Only

1.62

 indicates

1.62

 Taking

1.60

 Když

1.60

Activations Density 0.062%