INDEX

Explanations

repeated phrases and conjunctions that build coherence in a narrative

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

 simplement

-0.07

 simply

-0.07

çĲ³

-0.07

ãĥģãĥ¥

-0.06

Ð¸

-0.06

 Dabei

-0.06

 nackte

-0.06

ãģĸ

-0.06

æľīä»Ģä¹Ī

-0.06

 hopefully

-0.06

POSITIVE LOGITS

 during

0.15

 earlier

0.14

 recently

0.13

 last

0.12

 recent

0.11

during

0.11

recent

0.11

 During

0.11

 durante

0.10

 prior

0.10

Activations Density 0.089%