INDEX

Explanations

prime numbers greater than 1

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

 maxlen

-0.10

 Bran

-0.09

psilon

-0.08

IAM

-0.08

ounty

-0.08

Ð¸ÐµÐ¼

-0.08

 Essentials

-0.08

 MIME

-0.08

aph

-0.08

 ÑĥÑģÑĤÐ°Ð½Ð¾Ð²

-0.08

POSITIVE LOGITS

 prime

0.21

 Prime

0.19

Prime

0.19

prime

0.16

_prime

0.15

 primes

0.14

 composite

0.13

Composite

0.12

 itself

0.12

 Composite

0.12

Activations Density 0.076%