Model: SoLU Model: 8 Layers, 4096 Neurons per Layer

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

Neuron 1400 in Layer 2

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-8l')

Text #0

Max Range: 1.0523. Min Range: -1.0523

Max Act: 1.0523. Min Act: -0.0001

Data Index: 1060938 (C4 (Web Text))

Max Activating Token Index: 419

Text #1

Max Range: 1.0523. Min Range: -1.0523

Max Act: 1.0143. Min Act: -0.0001

Data Index: 496313 (C4 (Web Text))

Max Activating Token Index: 21

Text #2

Max Range: 1.0523. Min Range: -1.0523

Max Act: 0.9857. Min Act: -0.0001

Data Index: 1179035 (C4 (Web Text))

Max Activating Token Index: 878

Text #3

Max Range: 1.0523. Min Range: -1.0523

Max Act: 0.8972. Min Act: -0.0001

Data Index: 1184794 (C4 (Web Text))

Max Activating Token Index: 799

Text #4

Max Range: 1.0523. Min Range: -1.0523

Max Act: 0.8096. Min Act: -0.0001

Data Index: 50595 (C4 (Web Text))

Max Activating Token Index: 949

Text #5

Max Range: 1.0523. Min Range: -1.0523

Max Act: 0.8606. Min Act: -0.0001

Data Index: 357219 (C4 (Web Text))

Max Activating Token Index: 862

Text #6

Max Range: 1.0523. Min Range: -1.0523

Max Act: 0.9143. Min Act: -0.0002

Data Index: 793230 (C4 (Web Text))

Max Activating Token Index: 971

Text #7

Max Range: 1.0523. Min Range: -1.0523

Max Act: 0.9422. Min Act: -0.0002

Data Index: 154304 (C4 (Web Text))

Max Activating Token Index: 992

Text #8

Max Range: 1.0523. Min Range: -1.0523

Max Act: 0.9142. Min Act: -0.0001

Data Index: 202139 (C4 (Web Text))

Max Activating Token Index: 389

Text #9

Max Range: 1.0523. Min Range: -1.0523

Max Act: 0.9147. Min Act: -0.0001

Data Index: 1287786 (C4 (Web Text))

Max Activating Token Index: 935

Text #10

Max Range: 1.0523. Min Range: -1.0523

Max Act: 0.7723. Min Act: -0.0001

Data Index: 313097 (C4 (Web Text))

Max Activating Token Index: 483

Text #11

Max Range: 1.0523. Min Range: -1.0523

Max Act: 0.8718. Min Act: -0.0001

Data Index: 1598767 (Python Code)

Max Activating Token Index: 705

Text #12

Max Range: 1.0523. Min Range: -1.0523

Max Act: 0.8303. Min Act: -0.0001

Data Index: 1229131 (C4 (Web Text))

Max Activating Token Index: 496

Text #13

Max Range: 1.0523. Min Range: -1.0523

Max Act: 0.7974. Min Act: -0.0001

Data Index: 981257 (C4 (Web Text))

Max Activating Token Index: 942

Text #14

Max Range: 1.0523. Min Range: -1.0523

Max Act: 0.8394. Min Act: -0.0001

Data Index: 713369 (C4 (Web Text))

Max Activating Token Index: 749

Text #15

Max Range: 1.0523. Min Range: -1.0523

Max Act: 0.9043. Min Act: -0.0001

Data Index: 579745 (C4 (Web Text))

Max Activating Token Index: 659

Text #16

Max Range: 1.0523. Min Range: -1.0523

Max Act: 0.7977. Min Act: -0.0002

Data Index: 814321 (C4 (Web Text))

Max Activating Token Index: 608

Text #17

Max Range: 1.0523. Min Range: -1.0523

Max Act: 0.8237. Min Act: -0.0001

Data Index: 523252 (C4 (Web Text))

Max Activating Token Index: 850

Text #18

Max Range: 1.0523. Min Range: -1.0523

Max Act: 0.7436. Min Act: -0.0001

Data Index: 778322 (C4 (Web Text))

Max Activating Token Index: 708

Text #19

Max Range: 1.0523. Min Range: -1.0523

Max Act: 0.7729. Min Act: -0.0001

Data Index: 1301336 (C4 (Web Text))

Max Activating Token Index: 692