INDEX

Explanations

state of standing behind

The neuron strongly activates on words and phrases having to do with relative position or status—e.g. “standing,” “standings,” “behind,” “good standing,” “outstanding.”

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

 That

-1.43

to

-1.33

</h3>

-1.25

 there

-1.17

 тысячи

-1.13

 llegan

-1.10

-1.09

but

-1.08

as

-1.06

POSITIVE LOGITS

鱻

1.75

 preliminar

1.46

鱚

1.44

ഀ

1.36

 guarn

1.36

 adal

1.30

</b>

1.29

龗

1.26

塒

1.23

 custodia

1.23

Activations Density 0.002%