INDEX

Explanations

locations and specific entities

The neuron is primarily detecting numerical tokens (digits and number sequences) in the text.

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

 meille

-0.75

solved

-0.74

 trans

-0.73

Trans

-0.73

 Trans

-0.71

 Scottish

-0.71

Bey

-0.70

 program

-0.70

 bottoms

-0.70

Priv

-0.69

POSITIVE LOGITS

 أغسطس

0.85

ztus

0.78

assination

0.77

 سوف

0.76

ímp

0.76

 idolat

0.75

mleri

0.72

 subtype

0.72

CPF

0.70

ői

0.69

Activations Density 0.018%