INDEX

Explanations

programming constructs

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

.”

-2.27

",

-2.13

-2.11

});

-2.05

craper

-2.03

and

-1.99

";

-1.98

‘

-1.97

</h3>

-1.93

aneers

-1.88

POSITIVE LOGITS

 them

2.34

 these

2.28

 those

2.27

 decrease

1.79

⾞

1.78

 espoir

1.76

1.73

our

1.72

 perdre

1.67

焢

1.66

Activations Density 0.002%