INDEX

Explanations

rely on to do

The neuron fires on the infinitive marker “to” (i.e. instances of “to” introducing a verb).

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

to

-1.87

 these

-1.72

hasia

-1.52

-1.49

 semacam

-1.49

 such

-1.47

蹋

-1.41

 obligación

-1.41

 quella

-1.40

 vienas

-1.36

POSITIVE LOGITS

 reír

1.53

 deseja

1.52

ILERS

1.51

ctions

1.48

olden

1.47

()}>

1.46

 apesar

1.44

we

1.42

1.41

before

1.39

Activations Density 0.028%