INDEX

Explanations

variations of the word "sec" in different contexts, likely related to time or divisions

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ymoon

-0.08

ç¯ĩ

-0.07

 Reconstruction

-0.07

nze

-0.07

ATUS

-0.07

inspace

-0.07

NTAX

-0.07

zzo

-0.07

"":

-0.07

POSITIVE LOGITS

ret

0.18

und

0.14

RET

0.12

-ret

0.12

reta

0.12

ÑĢÐµÑĤ

0.12

unda

0.12

unden

0.11

ret

0.11

ÑĥÐ½Ð´

0.11

Activations Density 0.004%