INDEX

Explanations

professional knowledge and social concepts

The neuron detects abstract, often nominalized concepts or qualities (e.g. polarization, cohesion, expertise, mobilization) in the text.

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

 ersten

-1.70

 leeren

-1.44

 inizi

-1.32

 Still

-1.31

at

-1.29

still

-1.26

 pronti

-1.25

 отличается

-1.24

-1.23

 Similar

-1.22

POSITIVE LOGITS

 what

1.52

 this

1.32

 helicóptero

1.30

 chaves

1.28

 cinto

1.27

"]}

1.26

跶

1.25

 karta

1.24

UrlEncoded

1.24

 tela

1.23

Activations Density 0.037%