INDEX

Explanations

itself

The neuron fires on constructions where a noun is emphasized by being referred to “itself” (or “themselves”)—e.g. “the painting itself,” “the shop itself,” “the course itself,” etc.

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

in

-2.61

for

-1.85

up

-1.72

an

-1.67

 owing

-1.55

 since

-1.53

to

-1.53

-1.48

at

-1.47

POSITIVE LOGITS

 carreaux

1.71

䛗

1.70

 divino

1.60

 neumáticos

1.60

本身

1.59

fficking

1.58

 fundido

1.52

ciare

1.52

 decorados

1.52

 boîtes

1.50

Activations Density 0.012%