INDEX

Explanations

prepositions and suffixes

location and time indicators

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

 Below

0.41

↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵

0.40

↵↵↵

0.39

draw

0.38

↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵

0.38

 bell

0.38

↵↵↵↵↵↵↵

0.37

<unused63>

0.37

 वा

0.36

scroll

0.36

POSITIVE LOGITS

那里

0.44

 അവിടെ

0.42

前者

0.42

 dort

0.41

hydrocèle

0.41

 einfache

0.40

autre

0.39

 εκεί

0.39

 zusätzliche

0.39

jazd

0.39

Activations Density 0.000%