INDEX

Explanations

is believed to be

The neuron is triggered by occurrences of the verb “to be,” especially in contexts like “believed to be,” “is,” or “are.”

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

}{

-2.09

When

-1.84

}_{

-1.83

Tetapi

-1.73

 guerras

-1.64

名字

-1.62

地區

-1.62

又は

-1.62

Secara

-1.62

 misguided

-1.59

POSITIVE LOGITS

of

1.99

,;

1.77

鸲

1.73

蛱

1.59

 bedoeld

1.57

still

1.56

 omdat

1.55

section

1.53

鸸

1.50

 verschiedener

1.49

Activations Density 0.020%