INDEX

Explanations

make [positive quality]

The neuron detects positive promotional adjectives (and their associated intensifiers) that describe or praise an experience or offering.

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

 proč

-1.33

ucap

-1.30

 todėl

-1.28

ОВА

-1.23

čo

-1.19

beelden

-1.19

of

-1.18

óis

-1.17

 geschreven

-1.17

 pembuatan

-1.16

POSITIVE LOGITS

 той

1.62

 даже

1.54

 even

1.44

 même

1.43

ἵ

1.40

 same

1.37

 aunque

1.34

軋

1.34

 хотя

1.30

ratislava

1.29

Activations Density 0.042%