INDEX

Explanations

defining phrases

This neuron detects common English function words (e.g. “the,” “is,” “there,” “you”) that serve as basic grammatical or discourse markers.

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

-1.88

利率

-1.61

ardı

-1.58

儔

-1.45

or

-1.43

要么

-1.42

-1.41

 mães

-1.41

也只是

-1.41

 Орден

-1.40

POSITIVE LOGITS

are

1.94

 want

1.70

’

1.66

 prenda

1.60

 show

1.59

 different

1.54

map

1.54

 abiye

1.53

 проєкту

1.52

 silnika

1.52

Activations Density 0.008%