INDEX

Explanations

emphasizing importance or caution

New Auto-Interp

Configuration

Prompts (Dashboard)

238,145 prompts, 512 tokens each

Dataset (Dashboard)

lmsys + oasst1

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

suggestion

0.33

荐

0.32

 frying

0.32

 recommendation

0.31

 inroads

0.30

 transformation

0.30

就可以

0.29

 suggestions

0.29

 rigging

0.29

POSITIVE LOGITS

 beware

0.66

 consider

0.60

 помнить

0.58

 carefully

0.57

 heed

0.57

 always

0.54

 remember

0.54

 beachten

0.54

 ALWAYS

0.53

考慮

0.53

Activations Density 0.523%