INDEX

Explanations

Greek negation and verbs

Gemma team discussion

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

 einschließlich

0.46

噱

0.46

显著

0.45

 similit

0.43

 речо

0.43

 differentiated

0.42

 loopholes

0.42

 substantially

0.42

 conducción

0.42

 நுண்ண

0.41

POSITIVE LOGITS

 είναι

0.85

 έχ

0.82

 έχει

0.82

 βρί

0.79

 χρει

0.79

 μπορεί

0.78

 ήταν

0.78

έχ

0.77

 θα

0.75

 ξεκ

0.74

Activations Density 0.001%