INDEX

Explanations

field names like id, name, place

The neuron fires on the use of a dot after an identifier in code (i.e. object property or namespace access).

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

and

-1.84

 things

-1.78

 actionTypes

-1.41

 என்ப

-1.39

 those

-1.34

 something

-1.34

 inputStream

-1.33

즐

-1.28

 your

-1.28

 次元

-1.27

POSITIVE LOGITS

给我们

1.46

bledon

1.40

懨

1.39

housie

1.39

Their

1.38

 üblichen

1.37

 adecuado

1.34

给人

1.34

 любого

1.34

 निम्

1.34

Activations Density 0.030%