INDEX

Explanations

goods and services

The neuron activates specifically on mentions of “service” or “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

 need

-1.73

 Saltar

-1.52

 again

-1.52

}$

-1.51

and

-1.49

èche

-1.29

);

-1.25

北欧

-1.23

 zwar

-1.23

uramente

-1.23

POSITIVE LOGITS

 tabak

1.68

 akut

1.57

 which

1.46

 politika

1.46

 kamu

1.45

幛

1.42

 secos

1.35



1.35

 providers

1.34

 fotogra

1.34

Activations Density 0.010%