INDEX

Explanations

installing packages with pip

This neuron fires on numeric literals (especially floating-point numbers) in code or comments.

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

If

-1.70

 sehr

-1.63

麋

-1.55

на

-1.52

 troubling

-1.47

 indicates

-1.45

one

-1.45

 HUGE

-1.43

是的

-1.42

 sparkly

-1.41

POSITIVE LOGITS

the

1.94

1.86

%),

1.80

 vollen

1.79

_);

1.79

%).

1.77

嫋

1.72

1.71

葙

1.70

埽

1.70

Activations Density 0.004%