INDEX

Explanations

taking a walk

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

rap

-0.11

 vacation

-0.10

Rap

-0.10

 camping

-0.10

 Wilde

-0.09

 Camping

-0.09

 Pert

-0.09

ddl

-0.09

 Trav

-0.09

 Rush

-0.09

POSITIVE LOGITS

 walks

0.46

 walk

0.44

 walked

0.38

walk

0.36

 walking

0.36

 Walk

0.32

Walk

0.32

.walk

0.32

 Walking

0.29

_walk

0.29

Activations Density 0.083%