INDEX

Explanations

car sales registrations

The neuron fires on numeric and percentage figures (e.g. registration counts, decline rates, millions, thousands) in statistical or data‐heavy reports.

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 creating

-0.84

شاف

-0.81

 producing

-0.80

 providing

-0.75

screens

-0.75

£

-0.74

夫人

-0.73

 wüns

-0.72

 визна

-0.72

Creating

-0.71

POSITIVE LOGITS

 sales

1.56

 registrations

1.41

 deliveries

1.22

 sell

1.20

sales

1.17

销量

1.17

Sales

1.13

 sold

1.12

 Regist

1.10

regist

1.08

Activations Density 0.009%