INDEX

Explanations

device paths and commands

The neuron detects references to device file paths (especially entries under “/dev/”).

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

-1.59

-1.54

-1.45

-1.43

to

-1.42

-1.41

 собі

-1.40

of

-1.37

-1.34

POSITIVE LOGITS

 INTERESAR

1.57

เพราะ

1.53

 оригіналу

1.49

BeginContext

1.44

</sup>

1.41

mentaires

1.40

ขอบคุณ

1.40

Zapraszamy

1.39

 Aufenthalt

1.39

utenant

1.38

Activations Density 0.003%