INDEX

Explanations

the [specific noun]

This neuron detects numerical measurement tokens (e.g. decimal numbers and quantities) in scientific prose.

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

なども

-1.69

も含

-1.66

鹩

-1.62

については

-1.54

についても

-1.52

tront

-1.51

izability

-1.51

 ciudadanía

-1.50

燹

-1.50

 レザー

-1.48

POSITIVE LOGITS

 both

2.64

There

2.08

 Those

2.02

our

1.95

Both

1.91

 Both

1.90

 When

1.88

any

1.86

If

1.85

by

1.81

Activations Density 0.209%