INDEX

Explanations

emotional, legal, garage, chemical needs

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

èĪį

-0.09

 cush

-0.08

 Mills

-0.08

İ

-0.08

ickle

-0.08

ï¾ī

-0.08

 Alman

-0.08

 Nina

-0.08

erno

-0.08

POSITIVE LOGITS

 needs

0.57

 Needs

0.45

needs

0.43

Needs

0.37

 requirements

0.36

éľĢæ±Ĥ

0.31

éľĢè¦ģ

0.30

nhu

0.28

 need

0.28

 potrze

0.28

Activations Density 0.105%

No Known Activations