INDEX

Explanations

reaching the end

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

MÁS

-1.30

Packard

-1.20

 ALWAYS

-1.20

 kematian

-1.20

อื่น

-1.19

ének

-1.16

and

-1.16

 particolare

-1.16

of

-1.15

 straff

-1.15

POSITIVE LOGITS

最後まで

1.34

]$}

1.23

Abbreviations

1.13

aterdag

1.13

Referenties

1.12

ㄥ

1.06

 задание

1.06

↵

1.05

 tandis

1.05

hline

1.04

Activations Density 0.037%