INDEX

Explanations

financial cost

The main thing this neuron does is detect mentions of financial costs, prices, or monetary amounts.

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

 aguja

-1.46

 aktivi

-1.41

 résine

-1.35

ческих

-1.30

Филь

-1.27

因为它

-1.27

әрмәләр

-1.27

ógicos

-1.27

junctive

-1.27

#[

-1.26

POSITIVE LOGITS

to

1.93

 including

1.38

 cost

1.38

 costs

1.37

 while

1.32

 expenses

1.30

 mientras

1.25

 would

1.22

无尽

1.17

selho

1.16

Activations Density 0.051%