INDEX

Explanations

file stream variables

The neuron detects tokens involved in file I/O operations—especially file‐handle variables (e.g. “f”), open modes (e.g. “rb”), and “as”/“in” in file‐open constructs.

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

of

-2.11

都不知道

-1.38

下班

-1.28

躄

-1.28

 التالي

-1.26

だそうです

-1.24

from

-1.24

ons

-1.19

janos

-1.19

-1.18

POSITIVE LOGITS

గు

1.38

ἵ

1.38

網

1.34

innis

1.32

Hva

1.31

号線

1.29

Væ

1.28

اپ

1.28

succeeded

1.25

/*",

1.24

Activations Density 0.022%