INDEX

Explanations

that followed by verbs

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

estead

1.46

ayside

1.34

ogiri

1.27

ver

1.25

italization

1.24

凸

1.24

invest

1.21

ньо

1.20

 heading

1.19

 complete

1.17

POSITIVE LOGITS

ﻐ

2.23

And

2.18

Và

2.16

 וא

2.09

 وت

2.08

 অ্যান্ড

2.08

tę

2.06

端的

2.02

 alcuna

2.01

 وأ

2.01

Activations Density 0.033%