INDEX

Explanations

RESUMEN / Abstract / Sažetak

The neuron responds to standalone numeric tokens (e.g. volume, page, year, or identifier numbers).

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

ведите

-0.84

Import

-0.83

QUÉ

-0.82

ῳ

-0.82

 информация

-0.81

kräf

-0.79

ⅼ

-0.77

 utilisé

-0.76

haha

-0.75

 putih

-0.75

POSITIVE LOGITS

 received

1.08

 Received

0.91

Centro

0.85

 recieved

0.85

 attempts

0.82

 Institute

0.81

 receive

0.79

adomo

0.79

 attempt

0.78

 Triangle

0.78

Activations Density 0.034%