INDEX

Explanations

expressions of desire and suggestions for improvement

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

zav

-0.08

.override

-0.07

ifu

-0.07

overlay

-0.07

Ä©

-0.07

 zast

-0.07

strict

-0.07

ì©

-0.07

lesson

-0.07

ueba

-0.07

POSITIVE LOGITS

 ability

0.06

 addition

0.06

ulle

0.06

 abilities

0.06

 Kraft

0.06

 feature

0.05

 capability

0.05

UED

0.05

oren

0.05

Wah

0.05

Activations Density 0.011%