INDEX

Explanations

section headers

The neuron fires on the names of end‐of‐article sections (e.g. “References,” “External links,” “Bibliography,” “Notes,” “Footnotes,” 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

 keram

-1.57



-1.55

 psychiat

-1.55

亊

-1.48

 журнал

-1.42

 hemd

-1.41

média

-1.40

␮

-1.38

 múl

-1.38

 kabel

-1.38

POSITIVE LOGITS

达到了

1.59

 ursprüng

1.56

 предела

1.44

1.41

 такі

1.41

もいい

1.40

 に関する

1.38

這些

1.38

來了

1.37

絕對

1.34

Activations Density 0.010%