INDEX

Explanations

self-improvement and growth

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

opyright

-0.09

 beginning

-0.09

lish

-0.09

fÃ¼g

-0.09

ovi

-0.08

uppy

-0.08

zf

-0.08

lage

-0.08

lington

-0.08

ningar

-0.08

POSITIVE LOGITS

 improvement

0.41

 growth

0.41

 Improvement

0.36

 Growth

0.34

growth

0.31

Impro

0.28

 improve

0.27

 improvements

0.26

rowth

0.25

_growth

0.25

Activations Density 0.233%