INDEX

Explanations

import statements and header files

The neuron fires on tokens that name code modules or header files—e.g. the filenames (and their “.h” extensions) in #include or 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

-1.82

 natomiast

-1.48

 cães

-1.28

badah

-1.27

entière

-1.24

 многие

-1.24

鲻

-1.24

 некоторы

-1.20

cured

-1.20

 Condiciones

-1.19

POSITIVE LOGITS

it

1.47

its

1.38

它

1.37

Its

1.37

䛗

1.33

\%)

1.30

٣

1.28

Its

1.24

饜

1.23

 assister

1.20

Activations Density 0.033%