INDEX

Explanations

The neuron activates on the indefinite articles “a” and “an.”

a + role or status

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

 that

-1.82

any

-1.41

 specific

-1.34

 each

-1.27

 other

-1.24

 which

-1.24

 similar

-1.23

 others

-1.20

 different

-1.16

nikahan

-1.13

POSITIVE LOGITS

 also

1.51

of

1.43

 يحدث

1.30

 égard

1.27

 quí

1.27

 Kühlschrank

1.27

 princípios

1.26

 cóc

1.25

 авгу

1.24

très

1.23

Activations Density 0.057%