INDEX

Explanations

software tests

The neuron activates on testing-related identifiers and keywords (e.g. “Test,” “TestCase,” “assertTrue,” “@Test”).

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

тивным

-1.78

-1.72

 maravill

-1.60

 hermosos

-1.48

-1.44

————————————————

-1.38

曇

-1.38

럿

-1.38

ve

-1.37

POSITIVE LOGITS

<bos>

1.81

の詳細

1.56

淠

1.55

の概要

1.53

 ندار

1.45

 internacionais

1.45

ISTERS

1.44

alions

1.43

⢅

1.42

ного

1.41

Activations Density 0.002%