INDEX

Explanations

programming or technical utilities

The neuron activates on occurrences of the identifier “Utilities” (e.g. module, class, or file names) in code and documentation.

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

 aksesuar

-2.47

-2.45

 dekoratif

-2.42

en

-2.41

er

-2.38

wth

-2.38

we

-2.36

 Коли

-2.30

 avec

-2.23

POSITIVE LOGITS

?”

3.19

 preciosa

2.48

</sub>

2.38

镚

2.34

」

2.28

idjan

2.23

椟

2.20

2.17

逦

2.16

 mereka

2.13

Activations Density 0.002%