INDEX

Explanations

types and genres

The neuron fires on standalone metadata field or category labels (e.g. “make,” “format,” “type,” “color,” “capacity,” etc.).

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

!...

-1.07

orul

-1.05

さらに

-1.05

より

-1.03

‼

-0.98

 items

-0.96

цыя

-0.95

食べられる

-0.93

．．．

-0.93

谙

-0.93

POSITIVE LOGITS

 ندارد

1.52

Others

1.25

 VARIOUS

1.16

様々

1.16

doesn

1.16

depends

1.09

 None

1.09

 unclear

1.09

 OTHERS

1.06

seemed

1.05

Activations Density 0.099%