INDEX

Explanations

numbers and symbols

seperate cause or reason

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

ஃப்

0.53

—”

0.51

يًا

0.49

 কর্মকাণ্ড

0.49

ኅ

0.48

คาเดมี

0.48

ሠራ

0.48

َوْ

0.47

 애플

0.47

 FIXME

0.47

POSITIVE LOGITS

 ()=>{

1.16

={()=>

1.15

(()=>{

1.05

1.02

',()=>{

1.00

",(

0.99

()=>{

0.96

",&

0.91

',(

0.90

 ()=>

0.90

Activations Density 0.003%