Model: SoLU Model: 12 Layers, 6144 Neurons per Layer

Dataset: 80% C4 (Web Text) and 20% Python Code

Neuron 961 in Layer 9

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-12l')

Text #0

Max Range: 2.8598. Min Range: -2.8598

Max Act: 2.8598. Min Act: -0.0000

Data Index: 1584117 (Python Code)

Max Activating Token Index: 775

Text #1

Max Range: 2.8598. Min Range: -2.8598

Max Act: 2.8492. Min Act: -0.0000

Data Index: 1357752 (C4 (Web Text))

Max Activating Token Index: 380

Text #2

Max Range: 2.8598. Min Range: -2.8598

Max Act: 2.5408. Min Act: -0.0000

Data Index: 1388452 (Python Code)

Max Activating Token Index: 987

Text #3

Max Range: 2.8598. Min Range: -2.8598

Max Act: 2.5454. Min Act: -0.0000

Data Index: 746358 (C4 (Web Text))

Max Activating Token Index: 828

Text #4

Max Range: 2.8598. Min Range: -2.8598

Max Act: 2.6422. Min Act: -0.0000

Data Index: 319459 (C4 (Web Text))

Max Activating Token Index: 635

Text #5

Max Range: 2.8598. Min Range: -2.8598

Max Act: 2.4958. Min Act: -0.0000

Data Index: 823783 (C4 (Web Text))

Max Activating Token Index: 829

Text #6

Max Range: 2.8598. Min Range: -2.8598

Max Act: 2.5872. Min Act: -0.0000

Data Index: 1046706 (C4 (Web Text))

Max Activating Token Index: 994

Text #7

Max Range: 2.8598. Min Range: -2.8598

Max Act: 2.3572. Min Act: -0.0000

Data Index: 844128 (C4 (Web Text))

Max Activating Token Index: 201

Text #8

Max Range: 2.8598. Min Range: -2.8598

Max Act: 2.5741. Min Act: -0.0000

Data Index: 1445849 (Python Code)

Max Activating Token Index: 391

Text #9

Max Range: 2.8598. Min Range: -2.8598

Max Act: 2.4646. Min Act: -0.0000

Data Index: 436671 (C4 (Web Text))

Max Activating Token Index: 733

Text #10

Max Range: 2.8598. Min Range: -2.8598

Max Act: 2.4512. Min Act: -0.0000

Data Index: 1449701 (Python Code)

Max Activating Token Index: 572

Text #11

Max Range: 2.8598. Min Range: -2.8598

Max Act: 2.4429. Min Act: -0.0000

Data Index: 1471290 (Python Code)

Max Activating Token Index: 508

Text #12

Max Range: 2.8598. Min Range: -2.8598

Max Act: 2.5403. Min Act: -0.0000

Data Index: 1530622 (Python Code)

Max Activating Token Index: 843

Text #13

Max Range: 2.8598. Min Range: -2.8598

Max Act: 2.3887. Min Act: -0.0000

Data Index: 715700 (C4 (Web Text))

Max Activating Token Index: 805

Text #14

Max Range: 2.8598. Min Range: -2.8598

Max Act: 2.3540. Min Act: -0.0000

Data Index: 1570626 (Python Code)

Max Activating Token Index: 492

Text #15

Max Range: 2.8598. Min Range: -2.8598

Max Act: 2.3502. Min Act: -0.0000

Data Index: 858768 (C4 (Web Text))

Max Activating Token Index: 445

Text #16

Max Range: 2.8598. Min Range: -2.8598

Max Act: 2.2991. Min Act: -0.0000

Data Index: 1618713 (Python Code)

Max Activating Token Index: 978

Text #17

Max Range: 2.8598. Min Range: -2.8598

Max Act: 2.2153. Min Act: -0.0000

Data Index: 943238 (C4 (Web Text))

Max Activating Token Index: 272

Text #18

Max Range: 2.8598. Min Range: -2.8598

Max Act: 2.4184. Min Act: -0.0000

Data Index: 881450 (C4 (Web Text))

Max Activating Token Index: 656

Text #19

Max Range: 2.8598. Min Range: -2.8598

Max Act: 2.2959. Min Act: -0.0000

Data Index: 1186627 (C4 (Web Text))

Max Activating Token Index: 415