INDEX

Explanations

discussions related to the impact and evaluation of systems or practices

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

 gÃ¬

-0.06

it

-0.05

occo

-0.05

ex

-0.05

_integral

-0.05

 scand

-0.05

Jew

-0.05

 accord

-0.05

 Mellon

-0.05

POSITIVE LOGITS

0.09

unsch

0.08

aiser

0.08

Ã¡ty

0.07

utow

0.07

alone

0.07

_flutter

0.07

Ð²Ð¸ÑĩÐ°Ð¹

0.07

LK

0.07

azo

0.07

Activations Density 0.006%