INDEX

Explanations

focus and concentration

The main thing this neuron does is find mentions of mental focus and concentration.

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



-1.69

的身份

-1.66

 Officially

-1.64

 ANOTHER

-1.63

 Commonly

-1.61

骞

-1.60

巂

-1.60

 なく

-1.56

 странице

-1.55

的血

-1.55

POSITIVE LOGITS

 also

1.74

</h4>

1.65

The

1.60

and

1.59

on

1.48

 fhall

1.44

or

1.41

 create

1.39

），

1.34

 navnet

1.34

Activations Density 0.015%