INDEX

Explanations

tolerance and acceptable deviation

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

anita

-0.09

å¼ºåº¦é«ĺ

-0.08

æĴĳèµ·

-0.08

ustru

-0.08

 ê·ľëª¨

-0.08

uson

-0.07

rogen

-0.07

ìĵ¸

-0.07

itzer

-0.07

ThreadId

-0.07

POSITIVE LOGITS

 acceptable

0.14

åģıå·®

0.13

 toler

0.13

 tolerance

0.13

 deviation

0.12

acceptable

0.12

 allowable

0.11

 deviations

0.11

 tolerated

0.11

è¯¯å·®

0.11

Activations Density 0.024%