INDEX

Explanations

references to editorial notes or annotations

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

nt

-0.06

adopt

-0.06

uki

-0.06

lr

-0.06

if

-0.06

 ordinary

-0.05

ber

-0.05

 adoption

-0.05

asta

-0.05

asons

-0.05

POSITIVE LOGITS

@}

0.08

nist

0.08

icut

0.08

oload

0.08

aepernick

0.08

orial

0.07

Ã©ric

0.07

ials

0.07

'gc

0.07

 Chatt

0.07

Activations Density 0.001%