INDEX

Explanations

they

The neuron consistently lights up on numeric tokens and numerical data (e.g. years, statistics, reference numbers, measurements).

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

 వాటి

1.52

これらの

1.51

 these

1.45

这些

1.41

這些

1.41

 ඒවා

1.41

 সেগুলি

1.39

เหล่านี้

1.39

 அவற்றை

1.38

 этих

1.37

POSITIVE LOGITS

 انہوں

1.03

 उन्होंने

1.00

 చేశారు

1.00

 avevano

0.95

 کردند

0.93

उन्होंने

0.92

 использовали

0.92

 त्यांनी

0.89

 vieron

0.89

 encontraron

0.89

Activations Density 0.078%