INDEX

Explanations

each of X

This neuron detects instances of quantified counts or enumerations in the text (phrases like “each of N …” with numbers and item counts).

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

-2.05

er

-1.95

has

-1.91

min

-1.82

 Although

-1.76

 These

-1.72

In

-1.72

 When

-1.71

can

-1.70

二年

-1.70

POSITIVE LOGITS

⸫

2.19

熥

2.00

the

1.82

 najleps

1.80

盉

1.78

鼉

1.70

蕻

1.69

 mutta

1.68

koszulka

1.62

魨

1.61

Activations Density 0.012%