INDEX

Explanations

import statements and aliases

The neuron detects identifiers used as import aliases (e.g., the “as X” names in import statements).

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

at

-1.84

 only

-1.60

by

-1.56

for

-1.55

as

-1.48

on

-1.45

to

-1.43

in

-1.30

 more

-1.30

 Only

-1.27

POSITIVE LOGITS

我們

1.40

 Dinas

1.39

īts

1.35

ufc

1.32

totally

1.31

 två

1.30

 devast

1.28

:")

1.27

 instan

1.27

 sugger

1.26

Activations Density 0.006%