INDEX

Explanations

references to energy levels and stimulants

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

 hÃ¢l

-0.07

swer

-0.07

uppe

-0.07

_fatal

-0.07

_BAD

-0.07

 ÑĤÐµÐ¶

-0.07

icl

-0.06

 bazen

-0.06

ollipop

-0.06

fatal

-0.06

POSITIVE LOGITS

 overall

0.08

overall

0.07

quia

0.06

elez

0.06

 Overall

0.06

 normalized

0.06

 improvement

0.06

 ability

0.06

ILA

0.06

asto

0.06

Activations Density 0.014%