INDEX

Explanations

following tokens like next

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

Negative Logits

 prob

-0.11

ache

-0.11

 earlier

-0.11

uj

-0.09

 predecessors

-0.09

uo

-0.09

 Kann

-0.09

maal

-0.09

éĢĶ

-0.08

ught

-0.08

POSITIVE LOGITS

 next

0.66

next

0.52

_next

0.43

.next

0.43

Next

0.42

 nÃ¤chsten

0.41

 à¤ħà¤Ĺà¤²

0.40

-next

0.40

 Next

0.40

 nÃ¤ch

0.39

Activations Density 0.116%

No Known Activations