INDEX

Explanations

taking care of customers

multilingual translation and generation

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 betrayed

0.42

EP

0.41

asty

0.41

 intentionally

0.38

 fragmented

0.38

 December

0.37

 تھا۔

0.36

 Champion

0.36

 disorganized

0.36

ງ

0.35

POSITIVE LOGITS

 研究

0.49

 pokazuje

0.48

 soient

0.46

 aient

0.46

 acceso

0.46

 nắng

0.45

LEXPORT

0.45

gxh

0.45

 实现

0.44

 bezoek

0.44

Activations Density 0.000%