INDEX

Explanations

resignation and departure

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

ucks

-0.10

isco

-0.09

 extingu

-0.09

ØµØ¯

-0.09

 revival

-0.09

emmel

-0.09

gens

-0.09

 Closing

-0.08

aura

-0.08

 Hollow

-0.08

POSITIVE LOGITS

 resign

0.62

 resignation

0.57

 resigned

0.48

è¾ŀ

0.41

 quit

0.31

Res

0.30

 stepping

0.28

quit

0.28

Quit

0.27

 retirement

0.27

Activations Density 0.103%