INDEX

Explanations

The neuron strongly activates on four-digit numerals (e.g. years or similar large numeric codes).

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Europe

0.75

 એટ

0.72

ٹرول

0.71

من

0.68

 titular

0.67

MACl

0.67

europe

0.66

 اروپا

0.66

欧盟

0.65

 Eropa

0.64

POSITIVE LOGITS

스의

0.83

스

0.80

스로

0.80

 तब

0.77

스를

0.77

า

0.71

 ciclo

0.71

 серд

0.71

сний

0.71

сны

0.70

Activations Density 0.000%