INDEX

Explanations

head of

The neuron detects mentions of people in leadership roles introduced by “head of” (e.g., “head of” an organization or agency).

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

</td>

-1.59

 egyéb

-1.38

...'

-1.36

 setelan

-1.35

 سامسون

-1.35

 väldigt

-1.34

endalikan

-1.34

áček

-1.31

 geliefert

-1.30

wiada

-1.29

POSITIVE LOGITS

of

1.76

 still

1.38

its

1.30

 максима

1.30

 want

1.29

 continue

1.22

 their

1.20

不久

1.20

 reorgan

1.18

 overseeing

1.18

Activations Density 0.029%