INDEX

Explanations

change and alteration

The neuron strongly activates on the word “change” (and its variants like “changing” or “changes”), especially in contexts suggesting transformation or “change the world.”

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 প্রতিষ্ঠ

0.55

 achievements

0.53

زد

0.52

ত্যাগ

0.52

 Credentials

0.51

btnClose

0.50

出自

0.50

 logros

0.50

unc

0.50

 esfuerzo

0.50

POSITIVE LOGITS

改变

2.11

改變

2.06

 changed

2.03

 zmieni

1.95

 تغيير

1.88

changed

1.86

 измени

1.86

 föränd

1.86

 Changed

1.84

 cambiar

1.80

Activations Density 0.597%