INDEX

Explanations

week or month followed by day

The neuron activates on ordinal terms in date/time expressions (e.g. “last Sunday,” “second week,” etc.).

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

>\<

-0.85

 quarterly

-0.76

 declarado

-0.75

 каждому

-0.73

 olduk

-0.73

 concealing

-0.72

 suppressing

-0.71

 late

-0.71

pheny

-0.71

 Frühjahr

-0.71

POSITIVE LOGITS

整个

0.84

Rugby

0.83

мерикан

0.82

整個

0.81

 Fútbol

0.79

Џ

0.78

канал

0.77

 весь

0.76

entire

0.72

 entire

0.71

Activations Density 0.022%