Model: SoLU Model: 10 Layers, 5120 Neurons per Layer

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

Neuron 1240 in Layer 6

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-10l')

Text #0

Max Range: 2.3460. Min Range: -2.3460

Max Act: 2.3460. Min Act: -0.0000

Data Index: 715166 (C4 (Web Text))

Max Activating Token Index: 535

Text #1

Max Range: 2.3460. Min Range: -2.3460

Max Act: 2.2731. Min Act: -0.0000

Data Index: 680649 (C4 (Web Text))

Max Activating Token Index: 137

Text #2

Max Range: 2.3460. Min Range: -2.3460

Max Act: 2.1408. Min Act: -0.0000

Data Index: 1209696 (C4 (Web Text))

Max Activating Token Index: 738

Text #3

Max Range: 2.3460. Min Range: -2.3460

Max Act: 2.0775. Min Act: -0.0000

Data Index: 715414 (C4 (Web Text))

Max Activating Token Index: 683

Text #4

Max Range: 2.3460. Min Range: -2.3460

Max Act: 1.7668. Min Act: -0.0000

Data Index: 346431 (C4 (Web Text))

Max Activating Token Index: 1022

Text #5

Max Range: 2.3460. Min Range: -2.3460

Max Act: 1.8178. Min Act: -0.0000

Data Index: 1182244 (C4 (Web Text))

Max Activating Token Index: 85

Text #6

Max Range: 2.3460. Min Range: -2.3460

Max Act: 1.6991. Min Act: -0.0000

Data Index: 59508 (C4 (Web Text))

Max Activating Token Index: 869

Text #7

Max Range: 2.3460. Min Range: -2.3460

Max Act: 1.7588. Min Act: -0.0000

Data Index: 1539588 (Python Code)

Max Activating Token Index: 960

Text #8

Max Range: 2.3460. Min Range: -2.3460

Max Act: 1.7008. Min Act: -0.0000

Data Index: 1111345 (C4 (Web Text))

Max Activating Token Index: 451

Text #9

Max Range: 2.3460. Min Range: -2.3460

Max Act: 1.6751. Min Act: -0.0000

Data Index: 229459 (C4 (Web Text))

Max Activating Token Index: 189

Text #10

Max Range: 2.3460. Min Range: -2.3460

Max Act: 1.5616. Min Act: -0.0000

Data Index: 412360 (C4 (Web Text))

Max Activating Token Index: 753

Text #11

Max Range: 2.3460. Min Range: -2.3460

Max Act: 1.5692. Min Act: -0.0000

Data Index: 986813 (C4 (Web Text))

Max Activating Token Index: 996

Text #12

Max Range: 2.3460. Min Range: -2.3460

Max Act: 1.6285. Min Act: -0.0000

Data Index: 1141642 (C4 (Web Text))

Max Activating Token Index: 117

Text #13

Max Range: 2.3460. Min Range: -2.3460

Max Act: 1.5783. Min Act: -0.0000

Data Index: 92380 (C4 (Web Text))

Max Activating Token Index: 95

Text #14

Max Range: 2.3460. Min Range: -2.3460

Max Act: 1.6117. Min Act: -0.0000

Data Index: 1224579 (C4 (Web Text))

Max Activating Token Index: 740

Text #15

Max Range: 2.3460. Min Range: -2.3460

Max Act: 1.4706. Min Act: -0.0000

Data Index: 1259952 (C4 (Web Text))

Max Activating Token Index: 264

Text #16

Max Range: 2.3460. Min Range: -2.3460

Max Act: 1.6213. Min Act: -0.0000

Data Index: 762397 (C4 (Web Text))

Max Activating Token Index: 161

Text #17

Max Range: 2.3460. Min Range: -2.3460

Max Act: 1.4795. Min Act: -0.0000

Data Index: 1400759 (Python Code)

Max Activating Token Index: 603

Text #18

Max Range: 2.3460. Min Range: -2.3460

Max Act: 1.3405. Min Act: -0.0000

Data Index: 342489 (C4 (Web Text))

Max Activating Token Index: 428

Text #19

Max Range: 2.3460. Min Range: -2.3460

Max Act: 1.4531. Min Act: -0.0000

Data Index: 542837 (C4 (Web Text))

Max Activating Token Index: 83