INDEX

Explanations

in the

The neuron responds to content-bearing noun tokens (especially technical or domain-specific terms) rather than to common function words.

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

in

-2.66

ใน

-1.78

 into

-1.40

ในการ

-1.37

 στην

-1.36

 في

-1.33

在这

-1.23

在

-1.21

在我的

-1.19

在一个

-1.15

POSITIVE LOGITS

 וב

1.63

 وفي

1.32

 hinein

1.31

of

1.19

вом

1.14

 vicinity

1.12

 midst

1.10

看来

1.05

 doorway

1.01

guise

1.00

Activations Density 0.714%