INDEX

Explanations

dates, months

This neuron detects date‐time stamps, especially day‐of‐week abbreviations (e.g. “Mon,” “Tue,” “Wed,” “Thu,” “Fri”) in metadata lines.

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

oš

0.88

en

0.79

首先

0.72

boyfriend

0.71

第一次

0.69

初始

0.68

πό

0.68

当前

0.68

放

0.67

POSITIVE LOGITS

 microwaves

0.83

 wallpapers

0.81

 GaAs

0.81

 marbles

0.81

দিগকে

0.80

 profitieren

0.79

 Pergamon

0.79

Aud

0.78

 Sidi

0.78

 substituents

0.77

Activations Density 0.006%