INDEX

Explanations

hollows and indentations

The neuron detects the technical term “recess” (including its plural “recesses”).

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

 dimensione

-1.70

 manuten

-1.66

筇

-1.63

漚

-1.61

',

-1.61

筴

-1.55

 medes

-1.52

 适合

-1.52

 cilind

-1.52

 trasparente

-1.52

POSITIVE LOGITS

 depression

1.58

1.38

for

1.36

↵

1.35

1.32

1.31

 они

1.31

няют

1.30

жмите

1.27

all

1.23

Activations Density 0.098%