INDEX

Explanations

Q: : Punjab will Neural markers Answer one

The neuron activates on numeric tokens—particularly the volume, page, and year numbers found in academic citations and reference lists.

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

 teknik

-0.88

Pra

-0.87

 saman

-0.86

==""){

-0.85

 Tür

-0.85

 etik

-0.85

 merke

-0.84

DIRECT

-0.83

 teori

-0.83

 lavable

-0.82

POSITIVE LOGITS

 wüns

0.87

當然

0.86

producción

0.86

 özellikle

0.83

 located

0.82

 glaubte

0.81

 zdravot

0.80

どうぞ

0.79

 Asked

0.79

的時候

0.78

Activations Density 0.005%