INDEX

Explanations

body parts in specific positions

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

finger

-0.11

 finger

-0.10

til

-0.10

 tilted

-0.10

skirts

-0.10

Til

-0.10

Gow

-0.10

 fingert

-0.09

 hands

-0.09

vil

-0.09

POSITIVE LOGITS

 parallel

0.12

 straight

0.12

 extended

0.11

 overhead

0.11

 neutral

0.11

lee

0.11

 flex

0.11

ison

0.11

 abducted

0.11

extended

0.11

Activations Density 0.030%

No Known Activations