INDEX

Explanations

quantities like three or four

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

存储

1.10

 неу

1.09

 설계

1.07

中学校

1.05

 неско

1.04

 設計

1.03

唿

1.03

 వివా

1.03

 உற்பத்தி

1.02

 hazardous

1.01

POSITIVE LOGITS

 तीनों

1.82

all

1.52

いずれ

1.51

 كلهم

1.39

どれ

1.37

 trio

1.36

ALL

1.32

 trinity

1.31

 جميعا

1.31

 quartet

1.30

Activations Density 0.449%