INDEX

Explanations

westerly, northerly, Wester

The neuron is highly activated by tokens containing the substring “west” or “wester” (e.g. westerly, northwesterly, Southwestern).

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

-2.48

 propel

-1.83

The

-1.79

 продукти

-1.79

Any

-1.72

 большого

-1.71

 Wissenschaften

-1.70

 duże

-1.70

そうですね

-1.69

といった

-1.68

POSITIVE LOGITS

摃

2.64

焄

2.41

灩

2.31



2.28

鼴

2.22

儈

2.19

禩

2.13

彀

2.13

镒

2.11

 barrage

2.06

Activations Density 0.005%