INDEX

Explanations

Republic, Docket

The neuron fires on isolated multi-digit numeric tokens or long numeric/code sequences (e.g. document IDs, page numbers, or long digit strings).

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

-1.01

 ensamb

-1.00

 desse

-0.99

[]{"

-0.96

 août

-0.94

 teatral

-0.92

 tableta

-0.90

 jugend

-0.89

 lidi

-0.89

 Rusia

-0.88

POSITIVE LOGITS

vater

0.90

 qualcosa

0.88

 verantwort

0.87

 Уже

0.86

тный

0.82

推移

0.79

 derzeit

0.77

 conversely

0.77

 siguen

0.76

 болез

0.76

Activations Density 0.001%