INDEX

Explanations

through the means or outcome

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

 they

-2.02

ни

-2.00

 тот

-1.96

-1.93

enterOuterAlt

-1.91

ってました

-1.88

has

-1.85

was

-1.84

<>());

-1.83

 stället

-1.83

POSITIVE LOGITS

鹀

2.16

 danken

2.08

 становятся

2.06

 verlie

1.94

 fabricación

1.87

珦

1.86

 являются

1.85

 皇族

1.84

 排骨

1.84

 выполнять

1.83

Activations Density 0.056%