INDEX

Explanations

AI model behavior

section and list headers—signals of structured, enumerated or bulleted formatting in the text.

New Auto-Interp

Configuration

Prompts (Dashboard)

238,145 prompts, 512 tokens each

Dataset (Dashboard)

lmsys + oasst1

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 localised

0.28

 necrotic

0.27

 dens

0.27

 petits

0.26

 wastage

0.26

їз

0.25

 moist

0.25

 lagged

0.25

 immunost

0.25

 endings

0.25

POSITIVE LOGITS

John

0.31

Mike

0.30

Andrew

0.30

Ryan

0.30

Sarah

0.29

Why

0.29

Your

0.29

why

0.27

AI

0.27

 Ryan

0.27

Activations Density 0.184%