INDEX

Explanations

unit testing code

The neuron detects tokens related to testing—for example occurrences of “test,” “Test,” “Tests,” or names of testing tools/frameworks.

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

ült

-1.54

 teraz

-1.52

iniques

-1.45

 hemd

-1.43

 Teraz

-1.42

 اکنون

-1.42

 ehemal

-1.38

 vengono

-1.38

Сейчас

-1.38

 nachher

-1.36

POSITIVE LOGITS

<bos>

1.79

it

1.49

was

1.45

vább

1.42

<0xA3>

1.34

perity

1.31

ській

1.30

妄想

1.28

ле

1.27

 skóra

1.27

Activations Density 0.008%