INDEX

Explanations

services and what they include

The neuron fires on mentions of “service” or “services” (i.e. when the text is listing or referring to offered services).

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

 mostrare

-1.30

 from

-1.30

<bos>

-1.29

)}$,

-1.29

raktion

-1.27

cedure

-1.27

 time

-1.26

 predominant

-1.25

Արտաքին

-1.24

 getRule

-1.24

POSITIVE LOGITS

 include

1.46

یکه

1.45

 including

1.43

year

1.42

zied

1.35

 idag

1.34

Including

1.34

new

1.30

 fleuve

1.29

 banyo

1.29

Activations Density 0.037%