INDEX

Explanations

instances of the word "high" and related terms indicating a high standard or quality

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

 high

-0.48

high

-0.41

 High

-0.41

High

-0.39

-high

-0.36

 HIGH

-0.35

_high

-0.34

HIGH

-0.34

.high

-0.33

é«ĺ

-0.33

POSITIVE LOGITS

 LOWER

0.06

_LOWER

0.06

Lower

0.06

#undef

0.06

ppo

0.06

 Lower

0.05

urance

0.05

quette

0.05

abee

0.05

ansi

0.05

Activations Density 0.081%