INDEX

Explanations

something more or else

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

}}^{*}

2.06

iaire

2.06

ารา

1.93

}}^{*}\

1.80

 Debate

1.79

}^{*}$.

1.77

ioen

1.73

IBILITY

1.70

melee

1.69

ESTIC

1.69

POSITIVE LOGITS

 horribly

1.34

 unmistak

1.31

 strangely

1.30

似乎

1.28

偷偷

1.27

 inexplic

1.26

在这

1.25

始终

1.25

 somewhere

1.24

 непо

1.23

Activations Density 0.031%