INDEX

Explanations

later after time periods

This neuron detects numeric expressions of time intervals (e.g., a number paired with a time unit like days, weeks, years, decades).

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

-¹

0.60

はじ

0.52

}]_{

0.51

 ción

0.51

тивной

0.50

今年は

0.50

Ж

0.50

就要

0.50

ടുത

0.49

ট্ট

0.49

POSITIVE LOGITS

 later

3.02

later

2.72

 Later

2.67

Later

2.53

 später

2.27

 senere

2.23

 позже

2.22

 kasnije

2.13

 později

2.13

 senare

2.09

Activations Density 0.112%