INDEX

Explanations

situations involving play

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Stim

0.78

 resembled

0.73

->$

0.71

 resemble

0.70

 resembles

0.70

닦

0.69

 그러면

0.68

니아

0.67

 ontst

0.66

 확인함

0.65

POSITIVE LOGITS

 factored

1.48

 plays

1.43

 сыгра

1.41

 factor

1.36

 play

1.30

factor

1.29

 mattered

1.28

Factor

1.27

 played

1.27

因素

1.25

Activations Density 0.433%