INDEX

Explanations

configuration of a conventional

The neuron fires strongly on formal “example of…”‐style expressions in technical or patent writing.

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

ᾔ

-1.49

biť

-1.46

 ARISING

-1.38

toHaveBeenCalled

-1.32

zación

-1.30

 времени

-1.30

tawesome

-1.29

忑

-1.27

ripsi

-1.26

raborty

-1.26

POSITIVE LOGITS

our

1.83

for

1.72

 from

1.52

 veiks

1.45

1.34

the

1.34

at

1.31

how

1.27

of

1.23

one

1.20

Activations Density 0.051%