INDEX

Explanations

comment delimiters and two

The neuron is tuned to pick out tokens that are part of numeric or alphanumeric sequences—e.g. numbers, dates, measurements, addresses, citations.

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

多い

0.82

۲۰

0.80

્લા

0.79

۹

0.78

 محسوس

0.76

ammers

0.75

litre

0.75

 tentang

0.74

۲

0.73

流动

0.72

POSITIVE LOGITS

ка

0.72

га

0.71

עות

0.70

ел

0.69

ги

0.67

 ngủ

0.67

Easy

0.66

 excell

0.64

রোধ

0.64

Activations Density 0.000%