INDEX

Explanations

if not for / without

The neuron activates on standalone numeric tokens (integers or decimals).

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

is

-2.50

 When

-2.20

When

-1.80

has

-1.77

there

-1.70

膾

-1.68

when

-1.68

ensement

-1.64

 Когда

-1.63

In

-1.54

POSITIVE LOGITS

蕺

2.16

 those

2.06

 اللغة

1.96

泚

1.90

 โดย

1.89

᯾

1.86

 politika

1.83

 kablo

1.82

 また

1.80

쨩

1.80

Activations Density 0.003%