INDEX

Explanations

based on

The neuron strongly activates on the phrase “based on,” flagging instances where something is described as being founded or determined on some basis.

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

 moderne

-1.87

"""

-1.79

</td>

-1.72

’

-1.71

 judicia

-1.69

 normale

-1.66

 concr

-1.63

栳

-1.60

 russe

-1.54

kenen

-1.54

POSITIVE LOGITS

 that

1.78

 based

1.74

based

1.71

需要

1.62

 основан

1.55

.”.

1.55

 wichtigste

1.55

 focusing

1.47

 haupts

1.46

 basada

1.45

Activations Density 0.078%