INDEX

Explanations

time durations

This neuron detects numerical time-span expressions (e.g., “seven months,” “13 hours,” “two weeks”).

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

 vuoden

0.61

 Embar

0.59

 Recherches

0.57

 কয়েক

0.57

몇

0.57

 quelques

0.56

 временем

0.55

 Umum

0.55

 cuaderno

0.55

 remarqu

0.55

POSITIVE LOGITS

ของการ

0.76

 geleden

0.64

يم

0.62

ago

0.59

哚

0.56

-\

0.55

ه

0.55

of

0.54

=\

0.53

り

0.53

Activations Density 0.001%