INDEX

Explanations

this book analysis

The neuron detects mentions of content or publication types (e.g. “book,” “post,” “study”) indicating a work being referenced or reviewed.

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

 paździer

-1.63

dịch

-1.58

 Some

-1.56

Высота

-1.55

-1.53

 this

-1.52

',

-1.52

際には

-1.47

of

-1.45

not

-1.45

POSITIVE LOGITS

<bos>

2.08

Rptr

1.84

nique

1.81

ativement

1.80

"}";

1.72

孓

1.70

PropertyChanging

1.70

 tekrar

1.69

Verlag

1.68

sizeCache

1.66

Activations Density 0.187%