INDEX

Explanations

function context and ctx

The neuron strongly activates whenever the identifier “context” appears (e.g. as a function/constructor parameter or property).

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

’

-1.58

 grudnia

-1.45

’,

-1.37

 pomocou

-1.36

なければ

-1.33

 diariamente

-1.30

 colecciones

-1.30

 prič

-1.30

 comien

-1.28

存知

-1.27

POSITIVE LOGITS

 других

1.52

 russes

1.48

izer

1.47

嵛

1.44

 infierno

1.41

 الرئيسي

1.41

诠

1.41

mability

1.40

csolódó

1.40

 живут

1.39

Activations Density 0.004%