INDEX

Explanations

its roots, own, utility, allure

The neuron activates on the possessive pronoun “its.”

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

薷

-1.64

ارتمان

-1.55

 jedis

-1.55

 waifu

-1.51

cule

-1.51

蔟

-1.46

fordable

-1.46

樘

-1.45

仉

-1.43

vertising

-1.42

POSITIVE LOGITS

</h2>

3.19

</h4>

2.16

</strong>

1.78

$,

1.74

 also

1.66

 каждого

1.66

 каждой

1.63

1.57

 ทั้ง

1.57

 fueron

1.56

Activations Density 0.012%