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

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

Neuron 3139 in Layer 3

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.8527. Min Range: -0.8527

Max Act: 0.8527. Min Act: -0.0001

Data Index: 898689 (C4 (Web Text))

Max Activating Token Index: 614

Text #1

Max Range: 0.8527. Min Range: -0.8527

Max Act: 0.7391. Min Act: -0.0001

Data Index: 441198 (C4 (Web Text))

Max Activating Token Index: 627

Text #2

Max Range: 0.8527. Min Range: -0.8527

Max Act: 0.7399. Min Act: -0.0001

Data Index: 438770 (C4 (Web Text))

Max Activating Token Index: 246

Text #3

Max Range: 0.8527. Min Range: -0.8527

Max Act: 0.7353. Min Act: -0.0001

Data Index: 851048 (C4 (Web Text))

Max Activating Token Index: 194

Text #4

Max Range: 0.8527. Min Range: -0.8527

Max Act: 0.7242. Min Act: -0.0001

Data Index: 977504 (C4 (Web Text))

Max Activating Token Index: 638

Text #5

Max Range: 0.8527. Min Range: -0.8527

Max Act: 0.6995. Min Act: -0.0001

Data Index: 1031956 (C4 (Web Text))

Max Activating Token Index: 912

Text #6

Max Range: 0.8527. Min Range: -0.8527

Max Act: 0.7436. Min Act: -0.0001

Data Index: 1632080 (Python Code)

Max Activating Token Index: 780

Text #7

Max Range: 0.8527. Min Range: -0.8527

Max Act: 0.7226. Min Act: -0.0001

Data Index: 784697 (C4 (Web Text))

Max Activating Token Index: 390

Text #8

Max Range: 0.8527. Min Range: -0.8527

Max Act: 0.7009. Min Act: -0.0001

Data Index: 1045410 (C4 (Web Text))

Max Activating Token Index: 290

Text #9

Max Range: 0.8527. Min Range: -0.8527

Max Act: 0.7209. Min Act: -0.0001

Data Index: 1392894 (Python Code)

Max Activating Token Index: 1022

Text #10

Max Range: 0.8527. Min Range: -0.8527

Max Act: 0.6987. Min Act: -0.0001

Data Index: 1616340 (Python Code)

Max Activating Token Index: 649

Text #11

Max Range: 0.8527. Min Range: -0.8527

Max Act: 0.6946. Min Act: -0.0001

Data Index: 26435 (C4 (Web Text))

Max Activating Token Index: 9

Text #12

Max Range: 0.8527. Min Range: -0.8527

Max Act: 0.7027. Min Act: -0.0001

Data Index: 280613 (C4 (Web Text))

Max Activating Token Index: 298

Text #13

Max Range: 0.8527. Min Range: -0.8527

Max Act: 0.6603. Min Act: -0.0001

Data Index: 275744 (C4 (Web Text))

Max Activating Token Index: 879

Text #14

Max Range: 0.8527. Min Range: -0.8527

Max Act: 0.6690. Min Act: -0.0001

Data Index: 1551300 (Python Code)

Max Activating Token Index: 812

Text #15

Max Range: 0.8527. Min Range: -0.8527

Max Act: 0.6514. Min Act: -0.0001

Data Index: 252815 (C4 (Web Text))

Max Activating Token Index: 738

Text #16

Max Range: 0.8527. Min Range: -0.8527

Max Act: 0.6688. Min Act: -0.0001

Data Index: 391628 (C4 (Web Text))

Max Activating Token Index: 155

Text #17

Max Range: 0.8527. Min Range: -0.8527

Max Act: 0.6451. Min Act: -0.0001

Data Index: 66396 (C4 (Web Text))

Max Activating Token Index: 29

Text #18

Max Range: 0.8527. Min Range: -0.8527

Max Act: 0.6870. Min Act: -0.0001

Data Index: 185002 (C4 (Web Text))

Max Activating Token Index: 13

Text #19

Max Range: 0.8527. Min Range: -0.8527

Max Act: 0.6870. Min Act: -0.0001

Data Index: 103030 (C4 (Web Text))

Max Activating Token Index: 13