INDEX

Explanations

channel communications

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

 also

-2.25

栊

-2.20

秾

-2.06

 exprim

-1.98

 lendemain

-1.93

 Palabra

-1.87

観察

-1.87

 atau

-1.85

 débit

-1.84

 though

-1.84

POSITIVE LOGITS

इटम

2.55

</td>

2.50

<td>

2.00

 ""));

2.00

</b>

1.98

Usually

1.96

Often

1.93

1.92

 THEIR

1.91

ئم

1.91

Activations Density 0.004%