INDEX

Explanations

hypothetical "would" constructions

if you were

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

ங்களின்

0.41

 tenemos

0.40

 রয়েছে

0.39

돔

0.38

 fornis

0.37

ve

0.37

Pued

0.37

 admissions

0.36

 sólidos

0.36

开发的

0.36

POSITIVE LOGITS

would

0.79

 serait

0.67

 WOULD

0.66

 seraient

0.63

 would

0.61

 Would

0.61

 haría

0.58

zou

0.57

 würde

0.57

 করতাম

0.55

Activations Density 0.056%