INDEX

Explanations

conjunctions and their associated phrases relating to contrast or comparison

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

ninger

-0.06

Å¡i

-0.06

utz

-0.06

pij

-0.06

ÐµÐ·

-0.06

olin

-0.06

itably

-0.06

Vo

-0.06

essler

-0.06

ÑĮÑİ

-0.06

POSITIVE LOGITS

 less

0.09

 naopak

0.07

_LESS

0.07

but

0.07

lessness

0.07

 convers

0.07

 LESS

0.07

 fused

0.07

 moins

0.07

 Less

0.07

Activations Density 0.067%