INDEX

Explanations

showing research results

shows or demonstrates

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

嚷

0.45

 असतात

0.43

 असतो

0.41

 होंगे

0.39

 توضی

0.38

 হবেন

0.37

 będ

0.37

നാണ്

0.37

 रहेंगे

0.36

 ভেতরে

0.36

POSITIVE LOGITS

 shows

2.48

 showed

2.36

 показывает

2.30

shows

2.27

 menunjukkan

2.09

 pokazuje

2.08

 نشان

2.06

 montrent

2.06

 Shows

2.05

 zeigt

2.03

Activations Density 0.225%