INDEX

Explanations

one specific decision or event

The neuron activates on singular-quantity expressions—especially the word “one” (and synonyms like “single”) when used as a numeric or countable quantifier.

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

 uniquement

-1.43

actually

-1.25

 incess

-1.21

 plafon

-1.20

eventually

-1.18

presumably

-1.16

 olika

-1.15



-1.15

originally

-1.15

 maig

-1.13

POSITIVE LOGITS

or

1.79

one

1.70

 could

1.59

 might

1.34

may

1.23

 made

1.20

can

1.20

an

1.15

1.13

 changed

1.13

Activations Density 0.028%