INDEX

Explanations

objects and their properties

questions involving quantities or states

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

也都

0.41

也非常

0.38

还得

0.38

さすが

0.37

也不是

0.36

 إلي

0.36

也就

0.36

 travaillé

0.36

 Ahí

0.35

掌控

0.35

POSITIVE LOGITS

 undergoes

0.63

 whose

0.62

 जिसका

0.61

 weighs

0.60

whose

0.59

 absorbs

0.57

 cuya

0.57

 जिसकी

0.57

 loses

0.56

 lies

0.56

Activations Density 0.034%