INDEX

Explanations

possessive 's' followed by 's'

The neuron is detecting the possessive/genitive suffix "'s" appended to words.

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

笵

-1.70

 patatas

-1.65

纒

-1.63

不断的

-1.59

接下来的

-1.59

 古着

-1.57

 mundiales

-1.56

 bluse

-1.55

他会

-1.53

liski

-1.53

POSITIVE LOGITS

and

2.50

1.88

’

1.69

at

1.66

in

1.62

 avec

1.41

眺望

1.40

});*/

1.37

泊

1.34

illoin

1.34

Activations Density 0.069%