INDEX

Explanations

source of revenue

The main thing this neuron does is detect mentions of money, earnings, revenue, or other financial contexts.

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

’

-1.80

 also

-1.56

>}

-1.48

VIATIONS

-1.47

enment

-1.34

 there

-1.26

ituaries

-1.25

lcm

-1.23

と呼ば

-1.23

 referred

-1.23

POSITIVE LOGITS

corbic

1.58

cheiden

1.55

他的

1.50

of

1.47

「

1.45

denken

1.45

 serien

1.41

 музыки

1.40

睏

1.39

pergillus

1.37

Activations Density 0.017%