INDEX

Explanations

design

The neuron activates on occurrences of the substring “design” (e.g. “design,” “designed,” 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

for

-1.89

</i>

-1.52

 Occasionally

-1.40

 Immediately

-1.39

genheiten

-1.37

 Hundreds

-1.36

 Technically

-1.35

 Thousands

-1.34

 Equally

-1.32

^{\

-1.29

POSITIVE LOGITS

it

1.53

 แต่

1.51

 newest

1.41

豨

1.40

 bús

1.38

Asking

1.36

ाउन

1.35

 lack

1.34

 normaal

1.30

 calculer

1.28

Activations Density 0.033%