INDEX

Explanations

words and phrases indicating comparisons or contrasting elements

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

 indeed

-0.08

obec

-0.07

 exactly

-0.07

andon

-0.07

KeyId

-0.06

 takÃ©

-0.06

op

-0.06

ilent

-0.06

 Indeed

-0.06

abi

-0.06

POSITIVE LOGITS

 STILL

0.07

 still

0.07

 sometimes

0.07

azo

0.07

 succ

0.06

sson

0.06

eshire

0.06

 Ð¿Ð¾Ð±Ð°

0.06

enef

0.06

UILD

0.06

Activations Density 0.049%