INDEX

Explanations

initial or beginning

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

_tD

-0.09

imore

-0.09

_tE

-0.09

Statics

-0.09

ydk

-0.09

 Founded

-0.09

ãģĵãĤĵãģ«

-0.09

ismet

-0.08

addtogroup

-0.08

_tC

-0.08

POSITIVE LOGITS

 initially

1.01

 initial

0.71

 Initially

0.68

initial

0.66

Initially

0.62

æľĢåĪĿ

0.59

 inicial

0.57

 originally

0.49

 Initial

0.46

_initial

0.43

Activations Density 0.149%