INDEX

Explanations

held without charge or trial

New Auto-Interp

Top Features by Cosine Similarity

Configuration

Prompts (Dashboard)

10,000 prompts, 128 tokens each

Dataset (Dashboard)

lmsys/lmsys-chat-1m

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Enforcement

-0.11

 impeachment

-0.09

 jailed

-0.09

 enforcement

-0.09

 punishable

-0.09

å¯¸

-0.09

 incarcerated

-0.09

 Ãľl

-0.09

 losses

-0.09

aku

-0.09

POSITIVE LOGITS

 torture

0.26

 Tort

0.23

 TORT

0.19

 human

0.19

 Human

0.19

 tort

0.18

 habe

0.17

 mist

0.17

 tortured

0.16

Human

0.16

Activations Density 0.025%