INDEX

Explanations

for example

The neuron is highly sensitive to common English prepositions and other short function words (e.g. “in,” “for,” “by”).

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

</em>

-1.40

mathrm

-1.38

معرفی

-1.28

 imidlertid

-1.23

 dupa

-1.23

توضیحات

-1.22

 exquisite

-1.22

並不

-1.22

can

-1.20

POSITIVE LOGITS

 either

1.72

either

1.54

to

1.50

 isChecked

1.41

 entweder

1.30

Să

1.30

 unless

1.28

you

1.28

tram

1.27

たと

1.27

Activations Density 0.152%