INDEX

Explanations

judge names following `able`

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

“

-1.95

 gdyż

-1.94

-1.88

-1.79

 last

-1.78

 dejando

-1.76

 мужчин

-1.72

 nehmen

-1.70

 verwenden

-1.70

POSITIVE LOGITS

𖥧

2.08

蒽

1.94

 SEGU

1.89

Ꮉ

1.89

為に

1.86

 但是

1.81

⦿

1.76

⦊

1.74

際には

1.73

 TIEMPO

1.72

Activations Density 0.007%