INDEX

Explanations

who is the

The neuron is detecting non-English (foreign language) tokens—i.e. words or morphemes not in standard English.

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

ковой

-1.71

橚

-1.70

鲥

-1.69

</em>

-1.63

mientos

-1.62

 schwierig

-1.59

 Wortes

-1.55

 haupts

-1.51

 stables

-1.48

Ꮠ

-1.45

POSITIVE LOGITS

</h1>

2.39

</h2>

1.98

Such

1.85

âncias

1.77

 xvii

1.72

Enlla

1.71

ксана

1.68

<h2>

1.66

業

1.63

 artistico

1.61

Activations Density 0.011%