INDEX

Explanations

impedes, hampers, hinders

The neuron detects verbs (and their forms) that express obstruction or hindrance, such as “impede,” “hinder,” or “preclude.”

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

 öğretmen

-1.55

 lidé

-1.43

 llevaba

-1.41

未來

-1.37

坟

-1.34

去的

-1.34

disfraz

-1.34

بره

-1.34

 velkommen

-1.34

bisous

-1.34

POSITIVE LOGITS

2.44

</u>

2.06

2.00

="

1.82

</b>

1.81

1.73

1.71

耵

1.64

1.59

Activations Density 0.066%