INDEX

Explanations

stop followed by noun

The neuron strongly activates on requests or commands asking how to prevent or block something—especially the “stop … from …” construction.

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

を見た

-1.42

 equipamento

-1.35

 escuro

-1.33

女朋友

-1.32

-1.30

 imóvel

-1.28

 orçamento

-1.27

 already

-1.24

 then

-1.23

 devra

-1.23

POSITIVE LOGITS

gencias

1.70

by

1.62

1.56

);

1.47

?",

1.44

urably

1.41

嗵

1.37

id

1.34

任何

1.32

 industrie

1.31

Activations Density 0.029%