INDEX

Explanations

numbers and approximations

This neuron never activates on any token—it doesn’t appear to detect or fire for any pattern.

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

मिस्ट्री

0.58

মানে

0.57

বেশী

0.54

ruits

0.52

ार्मिक

0.52

étrico

0.52

uties

0.51

㭐

0.50

价格

0.50

рики

0.49

POSITIVE LOGITS

 około

0.82

 zehn

0.77

 aproximativ

0.77

 ۱۰

0.76

 ۲۰۰

0.74

 २००

0.71

 ۱۵

0.70

 aproximadamente

0.69

۶

0.68

0.66

Activations Density 0.000%