INDEX

Explanations

statements reflecting uncertainty or criticism about improvement and personal conduct

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

iky

-0.07

ngth

-0.07

leh

-0.06

ouncy

-0.06

utos

-0.06

inker

-0.06

leigh

-0.06

ä¼´

-0.06

atal

-0.06

áº¡ng

-0.06

POSITIVE LOGITS

 better

0.08

ä¹Łä¸į

0.08

 differently

0.08

 fare

0.08

 differs

0.07

 difference

0.07

 differ

0.07

better

0.07

 mieux

0.07

 difer

0.07

Activations Density 0.028%