INDEX

Explanations

code structure following `If` or `json`

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵

0.70

৪০

0.64

↵↵↵↵↵↵↵↵↵↵

0.63

↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵

0.61

↵↵↵↵↵↵↵↵

0.61

↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵

0.60

↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵

0.59

↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵

0.59

↵↵↵↵↵↵↵↵↵↵↵↵↵↵

0.57

↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵

0.57

POSITIVE LOGITS

 Similar

0.60

 Whale

0.59

 katt

0.58

Leave

0.57

ഞ്ഞ്

0.55

序

0.55

ér

0.54

 Recurs

0.54

Snippet

0.54

 panc

0.53

Activations Density 0.000%

No Known Activations

This feature has no known activations.