INDEX

Explanations

questions or phrases that seek clarification or detail

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

uck

-0.07

idis

-0.07

UCK

-0.07

 Yeni

-0.06

rupa

-0.06

uhe

-0.06

(reinterpret

-0.06

 hass

-0.05

ufs

-0.05

POSITIVE LOGITS

 many

0.16

 much

0.14

many

0.13

-many

0.12

 Many

0.11

 Much

0.11

 muchos

0.11

 viele

0.11

Many

0.11

much

0.11

Activations Density 0.021%