INDEX

Explanations

phrases indicating movement or transition

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

inction

-0.09

 *----------------------------------------------------------------

-0.08

OND

-0.08

odate

-0.08

peria

-0.08

cplusplus

-0.07

ipeg

-0.07

Dy

-0.07

LEncoder

-0.07

à¸ªà¸ģ

-0.07

POSITIVE LOGITS

 direction

0.07

 Direction

0.06

 angle

0.05

 Gomez

0.05

æĸ¹åĲĳ

0.05

 facets

0.05

 friction

0.05

 Secure

0.05

 foll

0.05

 hood

0.05

Activations Density 0.000%