INDEX

Explanations

toxic manipulation and abuse

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

 timelapse

0.41

 মৃত্যুদ

0.40

LinkClicked

0.39

 conject

0.39

 görsel

0.38

 discontinuities

0.38

 वान

0.38

 fichero

0.38

 irreducible

0.37

 satuan

0.37

POSITIVE LOGITS

 unhealthy

1.45

 toxic

1.37

 manipulative

1.28

Toxic

1.27

 abusive

1.20

 Toxic

1.18

 manipulation

1.16

 toxicity

1.16

toxic

1.16

 токси

1.16

Activations Density 0.149%