INDEX

Explanations

terms related to architectural styles, specifically modern and Art Nouveau

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

 Insets

-0.07

ImageContext

-0.06

ulton

-0.06

arness

-0.06

 Sands

-0.06

(KP

-0.06

ewan

-0.06

 Depths

-0.06

BaseContext

-0.06

abras

-0.06

POSITIVE LOGITS

Neo

0.13

neo

0.12

 Late

0.12

 Jugend

0.11

 neob

0.11

Bar

0.11

 Rational

0.11

ne

0.11

 Gothic

0.11

Neo

0.10

Activations Density 0.033%