INDEX

Explanations

now, currently, actually

The neuron is detecting multi-digit numeric tokens (large integer counts) in the text.

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

-2.56

-2.34

“

-2.09

ड

-1.89

is

-1.85

.《

-1.82

}$

-1.79

 anunció

-1.77

等人

-1.73

him

-1.73

POSITIVE LOGITS

and

2.13

别说

2.08

titial

2.06

خفض

1.95

之間

1.92

 prostřed

1.90

兩種

1.88

ennung

1.84

 می‌توان

1.82

しちゃ

1.81

Activations Density 0.081%