INDEX

Explanations

introduces clauses or facts

idea followed by that

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

 significantly

0.63

 والذي

0.43

qui

0.40

 이유는

0.39

有什么

0.38

さんは

0.37

 والتي

0.36

যার

0.36

 digamos

0.36

 economies

0.35

POSITIVE LOGITS

 bahwa

3.00

 rằng

2.64

 bahawa

2.63

 kwamba

2.53

 ότι

2.48

 that

2.33

że

2.33

 että

2.30

 أنه

2.28

that

2.25

Activations Density 1.402%