INDEX

Explanations

for paid surveys

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

agr

-0.10

oric

-0.09

discount

-0.09

 Marino

-0.09

ibi

-0.09

cdf

-0.08

 Ancient

-0.08

_ONCE

-0.08

 livest

-0.08

elman

-0.08

POSITIVE LOGITS

 opinion

0.17

 Opinion

0.17

 opin

0.14

 surveys

0.14

 earn

0.13

 survey

0.13

Op

0.13

 paid

0.13

 opinions

0.13

 rewards

0.12

Activations Density 0.045%