INDEX

Explanations

even followed by comparative

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

of

-1.80

at

-1.74

also

-1.70

 soixante

-1.69

 istimewa

-1.66

 décret

-1.52

 الأخرى

-1.50

者的

-1.45

 rechange

-1.45

Sixty

-1.45

POSITIVE LOGITS

<bos>

2.08

1.89

↵↵

1.63

</h1>

1.58

</th>

1.57

1.56

 than

1.48

 הוא

1.39

 worse

1.38

 more

1.33

Activations Density 0.012%