INDEX

Explanations

corporate announcements and governance

This neuron activates on the token “Corporate.”

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

嘰

-2.94

恹

-2.77

嘥

-2.73

ᚔ

-2.63

癰

-2.59

缡

-2.59

 Direito

-2.56

 VERY

-2.56

ются

-2.56

 decidedly

-2.50

POSITIVE LOGITS

3.17

 incessant

3.03

 subtly

3.00

 unrelenting

2.86

While

2.78

Through

2.77

In

2.75

ve

2.73

”،

2.69

 exquisite

2.61

Activations Density 0.005%