INDEX

Explanations

list manipulation and code constructs

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

ruk

2.47

тат

2.41

uem

2.31

खोरी

2.30

श्किल

2.30

zioni

2.29

рку

2.29

ھ

2.26

honneur

2.25

ಹಾಸ

2.24

POSITIVE LOGITS

UFACT

2.50

না

2.31

ান্ড

2.20

cro

2.19

िवारी

2.18

ться

2.18

eine

2.12

 drivers

2.10

 accused

2.10

 unsett

2.08

Activations Density 0.001%