INDEX

Explanations

mathematical formulas

The neuron responds to the last substantive token before a segment or sentence boundary, effectively flagging sentence-final words.

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

0.76

ed

0.69

CJ

0.69

elda

0.68

ar

0.66

Bos

0.66

 Arsen

0.66

arra

0.66

Carl

0.64

POSITIVE LOGITS

 Tujuan

0.81

 onde

0.80

 உண்

0.80

রুখ

0.79

 儿童

0.79

imbang

0.78

硋

0.78

гами

0.77

 tambour

0.77

 ifelse

0.77

Activations Density 0.000%