INDEX

Explanations

having said that

The neuron detects discourse-transition phrases that signal a qualification or counterpoint—e.g. “that said,” “having said that,” 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

-1.66

-1.48

-1.46

Both

-1.40

.”

-1.38

 from

-1.38

Regardless

-1.35

 when

-1.34

ließen

-1.33

Many

-1.31

POSITIVE LOGITS

 dezelfde

1.80

 bepaalde

1.80

of

1.76

establishing

1.60

 belangrijke

1.55

岽

1.54

潆

1.54

 mogelijkheden

1.52

妧

1.52

the

1.50

Activations Density 0.002%