INDEX

Explanations

specific terms related to measurement or precision

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

.scalablytyped

-0.08

VERTISE

-0.07

rega

-0.07

Ø±Ø³ÛĮ

-0.07

fad

-0.07

eos

-0.07

eus

-0.07

ovala

-0.07

adoo

-0.07

arella

-0.07

POSITIVE LOGITS

ight

0.24

 tight

0.20

IGHT

0.19

 Tight

0.18

tight

0.17

ightly

0.16

 tighter

0.15

ighter

0.15

igh

0.15

ite

0.14

Activations Density 0.004%