INDEX

Explanations

auxiliary verb 'do'

The neuron fires on occurrences of the auxiliary/main verb “do” (including its form “does”).

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

-2.00

of

-1.87

-1.65

他的

-1.57

 qualsiasi

-1.52

鹇

-1.51

たちが

-1.49

 geraten

-1.49

ходом

-1.49

-1.47

POSITIVE LOGITS

the

1.84

⽐

1.55

 według

1.53

."

1.42

娘

1.42

 сейчас

1.41

cją

1.41

tivamente

1.40

но

1.39

señor

1.38

Activations Density 0.014%