INDEX

Explanations

solving a problem or issue

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

âī¦

-0.11

ewe

-0.11

 Lester

-0.11

æķ

-0.10

umb

-0.10

eyer

-0.10

REFERRED

-0.09

lect

-0.09

arius

-0.09

POSITIVE LOGITS

 problem

0.29

 issue

0.22

 problema

0.20

problem

0.20

åķıé¡Į

0.19

 mystery

0.19

 problems

0.18

 Problem

0.18

 váº¥n

0.18

éĹ®é¢ĺ

0.18

Activations Density 0.064%