INDEX

Explanations

life of X and Y

The neuron fires on words that appear in biographical metadata or headings—especially “life,” “and,” “works,” “music,” “lives,” “died,” “birthday,” etc.—i.e. tokens signaling a biography‐style summary.

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

鸲

-1.64

踯

-1.63

のご紹介

-1.49

の紹介

-1.48

ícias

-1.48

atized

-1.46

鲞

-1.45

盉

-1.44

᯾

-1.42

TestingModule

-1.41

POSITIVE LOGITS

and

1.86

↵↵

1.52

 工業

1.25

These

1.23

 ingenious

1.17

“:

1.17

1.16

larımız

1.16

Another

1.15

勵

1.15

Activations Density 0.038%