INDEX

Explanations

get, gets, filter, models

The neuron flags dot‐separated attribute or method accesses (e.g. “models.objects.get”).

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

 scenery

-0.82

 submerged

-0.79

闷

-0.76

 Teste

-0.75

엽

-0.74

 violences

-0.73

テ

-0.73

ጊ

-0.71

 GARANT

-0.71

 hendes

-0.69

POSITIVE LOGITS

 filter

1.60

filter

1.54

 Filter

1.18

 FILTER

1.12

get

1.11

get

1.09

 filtering

1.05

FILTER

1.00

 gets

1.00

 Filtering

0.99

Activations Density 0.003%