INDEX

Explanations

comparisons involving the word "it" and references to improvement or change 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

itel

-0.06

oints

-0.06

issing

-0.06

ìĿ¸ì§Ģ

-0.06

 Clarkson

-0.06

idel

-0.05

 arose

-0.05

uba

-0.05

aji

-0.05

éĭ

-0.05

POSITIVE LOGITS

was

0.17

Was

0.16

_was

0.15

 been

0.15

was

0.15

Was

0.15

 Ð±ÑĭÐ»Ð°

0.14

 Ð±ÑĭÐ»

0.14

WAS

0.14

 Ð±ÑĥÐ»Ð°

0.13

Activations Density 0.042%