INDEX

Explanations

start of specific sections

The neuron is picking out formatting delimiters or markup markers (e.g. block‐comment “*/”, Markdown headings “'''”, 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

-2.02

-1.77

ارتمان

-1.76

-1.68

</i>

-1.67

ichtete

-1.64

-1.58

-1.53

≤

-1.46

MITTED

-1.46

POSITIVE LOGITS

적인

2.02

not

1.81

因为他

1.62

KURZ

1.55

žnost

1.54

 适合

1.53

不夠

1.52

Betyg

1.51

Välkommen

1.50

 bílé

1.48

Activations Density 0.000%