INDEX

Explanations

aka or other names, regions, or bases

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

 immense

-0.94

 fest

-0.89

 Kombat

-0.88

ENOMEM

-0.86

<.

-0.84

bedded

-0.84

oted

-0.84

ordinate

-0.83

ocarditis

-0.83

POSITIVE LOGITS

也就是

1.02

aka

0.97

<bos>

0.89

}]$

0.86

と呼ばれる

0.85

(=

0.84

AKA

0.83

СКИ

0.82

このような

0.82

 will

0.82

Activations Density 0.106%