INDEX

Explanations

between two options

This neuron detects explicit references to pairs or duality, especially words like “two” and “both” indicating exactly two items or choices.

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

大丈夫です

-1.59

 מאוד

-1.34

only

-1.30

そのような

-1.30

囿

-1.30

既に

-1.28

 heridos

-1.24

ditionally

-1.23

 even

-1.23

 only

-1.23

POSITIVE LOGITS

wać

1.48

izadores

1.38

 dikatakan

1.35

 should

1.34

adeira

1.34

۔

1.33

咤

1.32

1.31

}(

1.30

 italianos

1.30

Activations Density 0.060%