INDEX

Explanations

conclusions from facts

The neuron activates primarily on standalone numeric tokens (e.g., years, measurements, or other 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

 интерьера

-1.07

 veces

-1.07

 помещений

-1.01

 hyped

-1.00

 sienta

-1.00

 elige

-0.98

 кухни

-0.97

 empieza

-0.96

 салона

-0.96

||

-0.96

POSITIVE LOGITS

 therefore

1.32

 Therefore

1.31

 Поэтому

1.11

 поэтому

1.05

Therefore

1.05

Поэтому

1.02

so

0.98

 hence

0.97

——-

0.96

کنون

0.93

Activations Density 0.187%