INDEX

Explanations

regulation and changes

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

 ratified

-0.10

owitz

-0.10

 Dahl

-0.09

 Mask

-0.09

 mandate

-0.09

 Jury

-0.09

ationale

-0.09

 PLAN

-0.09

ington

-0.09

 mandates

-0.09

POSITIVE LOGITS

 regulation

0.19

 regul

0.18

 framework

0.17

 ÑĢÐµÐ³ÑĥÐ»Ð¸

0.17

framework

0.17

 regulatory

0.15

 rules

0.14

Reg

0.13

 regulate

0.13

 regulations

0.13

Activations Density 0.056%