phrases indicating a comparison or range

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Embeds

IFrame

Link

Negative Logits

‘

-0.06

CTIONS

-0.06

 Corpor

-0.06

 Recruitment

-0.06

ead

-0.06

 disregard

-0.05

 whilst

-0.05

 Corporation

-0.05

 Miche

-0.05

POSITIVE LOGITS

izmet

0.09

_$_

0.07

lays

0.07

nine

0.07

å®¿

0.07

ÑģÑĤÐ°Ð½

0.07

ocz

0.07

roti

0.07

ãĥ¼ãĤ¯

0.07

.toBe

0.07

Activations Density 0.000%

No Known Activations