INDEX

Explanations

multiple related computing entities

The neuron activates on expressions of quantity—numbers or words like “multiple” directly modifying technical plural nouns (e.g. databases, servers, processors, accounts).

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

 also

-1.51

ropriate

-1.29

++];

-1.26

ensively

-1.21

racuda

-1.20

 Pusat

-1.20

 ebenfalls

-1.20

тельности

-1.18

 part

-1.18

 jeszcze

-1.17

POSITIVE LOGITS

 there

1.41

 هناك

1.30

 begint

1.19

zdroj

1.14

 немає

1.14

خواهد

1.11

throughout

1.11

 such

1.09

orologio

1.06

 distintas

1.03

Activations Density 0.302%