INDEX

Explanations

The neuron fires on text that lists where a work has been published or is forthcoming—for example “stories have appeared (or are forthcoming) in ….”

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

 மிகப்பெரிய

0.67

 intenz

0.67

ܝ

0.66

 lésions

0.66

 vollständig

0.66

 cách

0.65

ল

0.65

 inicial

0.64

 કે

0.64

深圳市

0.63

POSITIVE LOGITS

 magazines

2.05

 magazine

1.98

magazine

1.91

 periodicals

1.87

杂志

1.85

 Magazine

1.84

Magazine

1.80

 newsletters

1.77

 Magazines

1.71

 newsletter

1.66

Activations Density 1.355%