INDEX

Explanations

challenges and difficulty

challenging things

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

䒺

0.37

ាប់

0.36

щено

0.35

 праців

0.35

 בני

0.35

pcm

0.35

 coef

0.34

 дум

0.34

疏

0.34

 сторо

0.34

POSITIVE LOGITS

 challenge

4.94

挑战

4.72

challenge

4.69

 Challenge

4.59

Challenge

4.53

挑戰

4.53

 CHALL

4.34

 challenges

4.19

challenges

4.09

 challenged

4.06

Activations Density 0.270%