INDEX

Explanations

medical abbreviations and terms

The neuron activates for proper nouns and specialized terminology, particularly recognizing names of people, places, scientific terms, and acronyms.

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

 that

-1.86

 باتری

-1.73

 parcs

-1.63

部分を

-1.61

 wied

-1.59

你去

-1.55

 inny

-1.55

詎

-1.55

eek

-1.53



-1.53

POSITIVE LOGITS

两个

1.84

 имеет

1.69

↵↵

1.66

 даже

1.65

之前

1.64

 ingenting

1.64

 lisäksi

1.50

现在

1.49

 mengadakan

1.48

 вони

1.45

Activations Density 0.202%