INDEX

Explanations

package names

The neuron strongly activates on the first identifier segment immediately following “com.” in Java `package` statements—that is, the vendor/org name in the package path.

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

を

-1.09

は

-0.97

？』

-0.95

が

-0.94

および

-0.91

籠

-0.90

のような

-0.89

CÓ

-0.87

vader

-0.86

Beskrivning

-0.85

POSITIVE LOGITS

/*-

1.20

になれば

1.06

のもの

1.05

perative

1.01

 altrett

0.98

なのだ

0.98

⑊

0.97

ome

0.94

idenz

0.92

づけ

0.89

Activations Density 0.196%