INDEX

Explanations

suitable for most

This neuron fires on phrases that state broad compatibility or suitability—e.g. “compatible with,” “suitable for,” “fits most,” “works with all,” 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

it

-1.77

Sedangkan

-1.55

Menurutnya

-1.45

Bahkan

-1.34

 about

-1.31

Tetapi

-1.30

tasche

-1.30

hows

-1.26

nadequate

-1.24

sheets

-1.23

POSITIVE LOGITS

and

1.45

 như

1.39

who

1.37

{\

1.30

of

1.29

 pozost

1.16

clientY

1.16

與

1.16

 buchen

1.16

You

1.16

Activations Density 0.074%