INDEX

Explanations

the word "Ob" in various contexts, likely focusing on observational or narrative statements

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

uspend

-0.09

ptions

-0.08

lun

-0.08

bull

-0.07

patrick

-0.07

lum

-0.07

)(((

-0.07

osate

-0.07

apple

-0.07

Ð¸Ð»Ñı

-0.07

POSITIVE LOGITS

ob

0.08

sets

0.07

rien

0.07

Ob

0.07

lique

0.07

(ob

0.06

oga

0.06

serve

0.06

ets

0.06

 Pink

0.06

Activations Density 0.008%