INDEX

Explanations

code conversion and definitions

The neuron fires on numeric literals appearing as arguments to the str(…) conversion function.

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

႐

-1.59

 takie

-1.49

sukienka

-1.45

 heard

-1.45

sightly

-1.44

IANE

-1.39

}^{\

-1.35

not

-1.34

preduce

-1.34

每

-1.33

POSITIVE LOGITS

of

2.03

er

1.84

 выглядит

1.68

 naszego

1.63

劇情

1.54

欧美

1.51

詳しくは

1.49

 interesan

1.48

これに

1.48

 שלנו

1.46

Activations Density 0.007%