INDEX

Explanations

numerical statistics and performance metrics in competitive contexts

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

à¸²à¸ĩ

-0.08

emm

-0.07

reesome

-0.06

amber

-0.06

haus

-0.06

 Hess

-0.06

repr

-0.06

oggles

-0.06

antity

-0.06

emens

-0.06

POSITIVE LOGITS

 ultimately

0.10

 Ultimately

0.09

 eventually

0.08

Ultimately

0.07

hu

0.07

ult

0.07

 Eventually

0.07

en

0.07

Eventually

0.06

enance

0.06

Activations Density 0.039%