INDEX

Explanations

instances of the word "position" in relation to numeric values

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

lore

-0.08

ogany

-0.08

mie

-0.07

essler

-0.07

apon

-0.07

é¾Ħ

-0.07

uya

-0.07

usal

-0.07

ends

-0.07

ansion

-0.07

POSITIVE LOGITS

nement

0.14

ally

0.13

ality

0.12

naires

0.11

als

0.10

naire

0.10

nal

0.09

nist

0.08

amiento

0.08

ioned

0.08

Activations Density 0.062%