INDEX

Explanations

self- actions/compounds

The neuron detects self-referential actions or processes, i.e. words or phrases expressing that something does or makes itself (e.g. “resolves itself,” “cooks itself,” “self-assembling”).

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

of

-1.78

щий

-1.28

 than

-1.27

 consiguiente

-1.23

cotta

-1.23

 melancho

-1.19

 ancak

-1.18

">“

-1.15

ждают

-1.14

and

-1.12

POSITIVE LOGITS

 durar

1.52

独特的

1.49

 caros

1.44

 viejas

1.43

 bonitos

1.43

 pago

1.39

 curar

1.39

 bonitas

1.38

垸

1.34

 victorias

1.33

Activations Density 0.032%