INDEX

Explanations

hereby granted

The neuron selectively activates on numeric tokens (e.g. standalone numbers, numeric literals, and digits in code).

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

jag

-0.75

 Creation

-0.73

мира

-0.71

Creation

-0.71

SafeArea

-0.69

Owen

-0.68

ßung

-0.68

jane

-0.68

 Invoice

-0.68

 Europe

-0.68

POSITIVE LOGITS

 nela

0.77

 Dolor

0.73

čiau

0.70

 Beleuchtung

0.68

 але

0.68

structs

0.68

(",

0.66

 Malabar

0.66

rels

0.66

 ties

0.64

Activations Density 0.018%