INDEX

Explanations

technical terms related to network models and performance metrics

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

.↵

-0.06

 thereby

-0.06

estr

-0.06

 Also

-0.06

 even

-0.06

 actually

-0.05

 expires

-0.05

ij

-0.05

POSITIVE LOGITS

 merupakan

0.10

æĺ¯æĪĳ

0.08

bson

0.08

ddf

0.07

æĺ¯ä¸Ģ

0.07

Å¡ov

0.07

 preval

0.07

HTTPHeader

0.07

koc

0.07

 widely

0.07

Activations Density 0.040%