INDEX

Explanations

The neuron fires on markup delimiters—primarily HTML tag tokens (e.g. `<p>`, `<em>`, `</li>`, etc.).

formatting characters like ., -, *, >

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

AÇÃO

-1.32

 kämp

-1.29

maining

-1.23

鬟

-1.19

fır

-1.14

OfWork

-1.13

 интерьере

-1.11

 lebih

-1.10

 cuma

-1.10

ței

-1.09

POSITIVE LOGITS

1.43

就可以了

1.43

↵

1.41

use

1.32

就行了

1.29

1.23

完成です

1.20

 zeigen

1.20

--

1.13

Activations Density 0.003%