INDEX

Explanations

first or only roles

The neuron activates on ordinal and ranking expressions—especially “first” (and similar terms like “second,” “only,” “fourth”) and their accompanying numeric or positional context.

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

 telas

-1.70

 itens

-1.50

 ilumin

-1.44

劻

-1.44

Ń

-1.44

|)

-1.42

}=\

-1.41

他们

-1.41

 encanto

-1.40

 anual

-1.38

POSITIVE LOGITS

 that

2.84

to

2.25

1.78

 zarówno

1.63

 praticamente

1.62

一起去

1.61

 fija

1.61

 cuyos

1.59

 than

1.58

of

1.54

Activations Density 0.072%