INDEX

Explanations

references to specific events or experiences over time

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

 discrepan

-0.08

eton

-0.08

.MixedReality

-0.08

esson

-0.08

è©

-0.07

egas

-0.07

æľĢæĸ°

-0.07

_ONCE

-0.07

_TRNS

-0.07

retry

-0.07

POSITIVE LOGITS

pre

0.10

Pre

0.09

 already

0.08

 previous

0.07

 preliminary

0.07

 smaller

0.07

 informal

0.07

 Prel

0.07

 preview

0.07

 intermediate

0.07

Activations Density 0.065%