INDEX

Explanations

instructions and recommendations for planning and preparation

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

entic

-0.06

 because

-0.06

andard

-0.06

 itself

-0.06

ref

-0.06

enter

-0.06

because

-0.06

oret

-0.06

 thanks

-0.06

POSITIVE LOGITS

 accordingly

0.12

braco

0.08

/inet

0.08

ç°

0.08

iffe

0.07

 CONSEQUENTIAL

0.07

odzi

0.07

 Accordingly

0.07

ìĭ¬

0.07

ÙĪØ§

0.07

Activations Density 0.027%