INDEX

Explanations

structures and objects

The neuron responds strongly to concrete nouns denoting physical objects or structures (e.g. “circle,” “ladder,” “bucket,” “billboard,” “platform”).

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

-1.47

 asamblea

-1.46

 reizen

-1.41

 osoba

-1.38

 heilig

-1.36

 aplicativos

-1.35

 francus

-1.35

 keram

-1.34

änderung

-1.33

melons

-1.33

POSITIVE LOGITS

</i>

1.33

of

1.32

 construite

1.23

 vacía

1.16

%;

1.15

zik

1.14

||

1.14

 favorita

1.13

out

1.12

data

1.12

Activations Density 0.258%