INDEX

Explanations

helpful and harmless

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

з

1.13

"(

1.08

 его

1.06

have

1.05

 সকল

1.04

 Они

1.04

 प्रत्येक

1.03

((

1.03

has

1.03

equal

1.02

POSITIVE LOGITS

 übrigens

2.76

 however

2.45

 toutefois

2.06

 ovviamente

2.06

 évidemment

2.04

 anyway

2.00

 meanwhile

1.99

 obviamente

1.97

 oczywiście

1.97

 кстати

1.97

Activations Density 0.808%