INDEX

Explanations

references to tasks and responsibilities related to work or processing

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

ieber

-0.08

adel

-0.07

orton

-0.07

iba

-0.06

çŀ

-0.06

urge

-0.06

ECTOR

-0.06

OLF

-0.06

 Practices

-0.06

urga

-0.06

POSITIVE LOGITS

 heavy

0.15

 grunt

0.14

leg

0.13

 dirty

0.13

heavy

0.13

 Heavy

0.12

dirty

0.12

 tedious

0.11

Heavy

0.11

 behind

0.10

Activations Density 0.038%