INDEX

Explanations

arrow/drag/self/DB/HTTP/xmlns

The neuron strongly activates on HTML tag names (as seen in the comma-separated list of elements in a CSS reset).

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

 with

-1.55

at

-1.43

and

-1.41

its

-1.38

 their

-1.35

 bicicletas

-1.30

in

-1.23

 without

-1.20

 from

-1.19

one

-1.19

POSITIVE LOGITS

這些

1.16

 Además

1.15

 dando

1.11

 moron

1.07

これらの

1.07

oon

1.06

 afore

1.05

这些

1.04

 綺麗

1.03

 ameste

1.03

Activations Density 0.035%