INDEX

Explanations

training time for models

The main thing this neuron does is detect numeric tokens (especially floating-point and integer strings) in reference citations and bibliographic metadata.

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

िख

0.70

0.68

0.66

 corona

0.65

corona

0.65

most

0.64

conscious

0.64

messe

0.63

蔦

0.63

voucher

0.63

POSITIVE LOGITS

 ケース

0.94

 Diakses

0.92

 Agregar

0.86

 actúa

0.85

 Demonstrated

0.85

 Deshalb

0.84

 थी

0.84

 desenvolver

0.84

 وكان

0.83

 オフィス

0.83

Activations Density 0.000%