INDEX

Explanations

No Explanations Found

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

 Prefer

0.97

 Clearing

0.96

 IRAS

0.90

 Comptroller

0.89

 indemnity

0.89

custId

0.87

 Accounting

0.87

 Monitoring

0.86

AIA

0.85

 demarcation

0.85

POSITIVE LOGITS

 reworked

1.52

 transformed

1.48

 reimag

1.45

 repurposed

1.45

 reinterpret

1.39

变成了

1.38

transformed

1.37

 trasform

1.37

改造

1.33

 modified

1.30

Activations Density 0.780%

No Known Activations