INDEX

Explanations

failed attempts or unexpected outcomes

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

 ['*.

1.25

 কর্তৃক

1.16

 နှင့်

1.07

}_{{\

1.07

 기업

1.04

 thúc

1.04

 Specifies

1.03

 trillions

1.03

 môžu

1.02

 জনগোষ্ঠীর

1.01

POSITIVE LOGITS

had

1.47

感觉

1.32

Had

1.31

すごく

1.30

めっちゃ

1.24

 hice

1.24

やっぱり

1.24

良かった

1.22

すごい

1.22

had

1.21

Activations Density 0.926%