INDEX

Explanations

to and what

The neuron is strongly triggered by short function words that signal direction or movement—especially the preposition “to” (and similar tokens like “up” or “within”).

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

 bleak

-1.44

 wonderfully

-1.41

 dando

-1.40

 amazingly

-1.35

 serene

-1.34

 REALLY

-1.34

 breathtaking

-1.31



-1.31

 acht

-1.30

또

-1.30

POSITIVE LOGITS

。

2.83

 because

2.59

։

2.03

。「

2.00

।

1.93

.\\

1.92

1.90

}$.

1.85

 when

1.85

1.83

Activations Density 0.432%