INDEX

Explanations

head tightly

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

.scalablytyped

-0.13

hug

-0.10

erial

-0.09

 Bella

-0.09

 Witness

-0.09

rint

-0.09

 Lena

-0.09

 curled

-0.09

 inward

-0.09

ent

-0.08

POSITIVE LOGITS

/head

0.14

 tilt

0.13

 chin

0.13

 head

0.13

-head

0.12

(head

0.12

 nods

0.12

til

0.12

 nodded

0.11

gear

0.11

Activations Density 0.019%