INDEX

Explanations

she drank, is injected

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

 powdered

-0.12

 hydrated

-0.11

 Beer

-0.11

 beers

-0.11

dry

-0.11

 dust

-0.11

 washed

-0.10

 beer

-0.10

 Champagne

-0.10

gum

-0.10

POSITIVE LOGITS

 bottle

0.23

 liquid

0.22

 pour

0.21

 poured

0.21

 Pour

0.20

pour

0.20

æ¶²

0.20

 bottles

0.19

Pour

0.19

 fluid

0.19

Activations Density 0.100%