INDEX

Explanations

indmed

The neuron is responding to long alphanumeric identifier tokens—e.g. file or image names, URLs, or codes packed with letters and especially digits.

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

 hadas

-1.55

espejo

-1.52

 сына

-1.51

ᵴ

-1.49

TIPOS

-1.49

тории

-1.47



-1.46

aktionen

-1.44

 capucha

-1.43

 llavero

-1.42

POSITIVE LOGITS

in

1.63

we

1.58

of

1.51

 also

1.30

<h3>

1.27

 хвилин

1.27

1.25

 Each

1.24

 some

1.23

in

1.20

Activations Density 0.037%