INDEX

Explanations

private method definitions

The neuron selectively fires on salient content words—especially descriptive adjectives and key nouns—used in promotional, metadata, or image‐description contexts.

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

public

0.66

PUBLIC

0.61

angular

0.57

ങ്ങളിലും

0.56

Public

0.55

 ঘোষিত

0.55

NgModule

0.54

要望

0.54

ción

0.54

otype

0.54

POSITIVE LOGITS

 private

1.52

private

1.34

 Private

1.33

 PRIVATE

1.24

 privately

1.23

 privados

1.21

 निजी

1.19

私

1.18

Private

1.16

 privat

1.14

Activations Density 0.010%