INDEX

Explanations

code syntax formatting

This neuron detects template‐expression delimiters and other multi‐character punctuation tokens (e.g. backticks and mustache‐style `{{ … }}` markers).

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

次的

-0.81

norr

-0.79

داً

-0.77

Yah

-0.73

 кра

-0.73

 commercial

-0.71

 blind

-0.69

 complexion

-0.68

 intertw

-0.67

司

-0.67

POSITIVE LOGITS

“

0.84

...)

0.79

'#'

0.72

रीर

0.71

}},

0.71

PAI

0.71

 inicio

0.70

 LASER

0.69

sition

0.69

roscopy

0.69

Activations Density 0.118%