INDEX

Explanations

work closely with

FlashAttention, plum blossoms

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

0.50

 hacerse

0.43

ה

0.43

кус

0.43

 inlets

0.42

如此

0.42

εργ

0.41

ン

0.41

NCIA

0.40

سا

0.40

POSITIVE LOGITS



0.57

ʙ

0.57

くない

0.55

 पासवर्ड

0.54

 сотрудников

0.53

 مستقیم

0.53

 pracowników

0.52

 बँके

0.51

 सश

0.50

 uczniów

0.49

Activations Density 0.000%