INDEX

Explanations

understanding and correctness

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

षण

1.12

 inicio

0.99

続け

0.95

 inició

0.90

 resultant

0.88

 Inicio

0.87

 positives

0.85

 attributable

0.85

 прият

0.85

鞭

0.85

POSITIVE LOGITS

 rethinking

1.42

 thinking

1.27

 Paradigm

1.26

thinking

1.25

思维

1.22

 paradigm

1.20

 rethink

1.19

 paradigms

1.19

 Thinking

1.14

 pondering

1.13

Activations Density 0.266%