INDEX

Explanations

phrases indicating effort and the challenges of overcoming obstacles

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

uche

-0.07

wij

-0.06

echa

-0.06

ÑĢÐ°Ð²Ð°

-0.06

eyer

-0.06

/we

-0.06

ehr

-0.06

cky

-0.06

Ð»Ð¸ÑĤ

-0.06

rtl

-0.06

POSITIVE LOGITS

 necessity

0.09

 because

0.09

 choice

0.08

 Reasons

0.08

 reasons

0.08

because

0.08

 habit

0.07

 partly

0.07

 lack

0.07

 principle

0.07

Activations Density 0.044%