INDEX

Explanations

professional works and creative outputs

The neuron fires on words naming a person’s professional or creative output (e.g. “research,” “work,” “writing,” “music,” “artworks,” 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

 with

-1.51

 from

-1.25

 when

-1.18

to

-1.16

 should

-1.15

in

-1.13

or

-1.11

as

-1.05

 without

-1.04

by

-1.02

POSITIVE LOGITS

 range

2.17

 ranged

1.98

 ranging

1.92

ranging

1.85

 span

1.85

 spans

1.82

 ranges

1.75

covering

1.72

 cover

1.69

 covers

1.64

Activations Density 0.087%