INDEX

Explanations

ending or conclusion

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 intake

-0.11

 Intro

-0.10

 Badge

-0.10

-ip

-0.10

 intimidated

-0.09

 intox

-0.09

 Pause

-0.09

flu

-0.09

pause

-0.09

 pauses

-0.09

POSITIVE LOGITS

 ending

0.34

 endings

0.31

 Ending

0.26

 finale

0.25

Ending

0.25

 resolution

0.24

 concluding

0.24

 final

0.23

-ending

0.23

 conclusion

0.22

Activations Density 0.225%