INDEX

Explanations

version of

The neuron activates on phrases describing one work as a particular “version,” “adaptation,” or “extract” of another.

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

-1.77

 όμως

-1.70

boneca

-1.62

</h4>

-1.60

 there

-1.54

ござい

-1.49

<bos>

-1.48

fusions

-1.47

takings

-1.46

appartamento

-1.45

POSITIVE LOGITS

–

1.95

the

1.70

 what

1.66

鹇

1.52

jednoc

1.52

】：

1.50

 gerichtet

1.48

鯓

1.44

vn

1.42

backslash

1.41

Activations Density 0.099%