INDEX

Explanations

numbers or count

The neuron is keyed to prominent event-reporting verbs—words like “launched,” “announced,” “filed,” and “embarked” that mark major actions or newsworthy events.

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

そういう

1.09

 nějak

0.94

 irgend

0.90

 solche

0.88

 andere

0.85

 solchen

0.82

 다른

0.82

他の

0.81

 sowas

0.81

こういう

0.81

POSITIVE LOGITS

its

1.55

 نخست

1.17

 their

1.16

 ensimmä

1.16

two

1.13

 four

1.11

 three

1.10

 unprecedented

1.07

初の

1.06

 δύο

1.05

Activations Density 0.689%