INDEX

Explanations

references to research methods and techniques in the context of scientific studies

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

memberof

-0.09

jeme

-0.08

eing

-0.08

itere

-0.08

licity

-0.08

áº³ng

-0.08

_icons

-0.08

Disposition

-0.07

éĿ©

-0.07

ebo

-0.07

POSITIVE LOGITS

 learning

0.09

 Learning

0.08

learn

0.08

 learn

0.07

-learning

0.07

learning

0.06

åŃ¦ä¹ł

0.06

Learn

0.06

 retaining

0.06

 retained

0.06

Activations Density 0.006%