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

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

Neuron 3160 in Layer 8

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 1.1249. Min Range: -1.1249

Max Act: 1.1249. Min Act: -0.0000

Data Index: 1604603 (Python Code)

Max Activating Token Index: 220

Text #1

Max Range: 1.1249. Min Range: -1.1249

Max Act: 0.7936. Min Act: -0.0000

Data Index: 760084 (C4 (Web Text))

Max Activating Token Index: 458

Text #2

Max Range: 1.1249. Min Range: -1.1249

Max Act: 0.7685. Min Act: -0.0000

Data Index: 458687 (C4 (Web Text))

Max Activating Token Index: 827

Text #3

Max Range: 1.1249. Min Range: -1.1249

Max Act: 0.7258. Min Act: -0.0000

Data Index: 441995 (C4 (Web Text))

Max Activating Token Index: 188

Text #4

Max Range: 1.1249. Min Range: -1.1249

Max Act: 0.6656. Min Act: -0.0000

Data Index: 1270393 (C4 (Web Text))

Max Activating Token Index: 89

Text #5

Max Range: 1.1249. Min Range: -1.1249

Max Act: 0.6457. Min Act: -0.0000

Data Index: 785762 (C4 (Web Text))

Max Activating Token Index: 894

Text #6

Max Range: 1.1249. Min Range: -1.1249

Max Act: 0.6485. Min Act: -0.0000

Data Index: 884965 (C4 (Web Text))

Max Activating Token Index: 395

Text #7

Max Range: 1.1249. Min Range: -1.1249

Max Act: 0.6408. Min Act: -0.0000

Data Index: 1362421 (Python Code)

Max Activating Token Index: 698

Text #8

Max Range: 1.1249. Min Range: -1.1249

Max Act: 0.6301. Min Act: -0.0000

Data Index: 543063 (C4 (Web Text))

Max Activating Token Index: 585

Text #9

Max Range: 1.1249. Min Range: -1.1249

Max Act: 0.5806. Min Act: -0.0000

Data Index: 1370574 (Python Code)

Max Activating Token Index: 531

Text #10

Max Range: 1.1249. Min Range: -1.1249

Max Act: 0.6081. Min Act: -0.0000

Data Index: 1486452 (Python Code)

Max Activating Token Index: 602

Text #11

Max Range: 1.1249. Min Range: -1.1249

Max Act: 0.6070. Min Act: -0.0001

Data Index: 1622003 (Python Code)

Max Activating Token Index: 94

Text #12

Max Range: 1.1249. Min Range: -1.1249

Max Act: 0.5787. Min Act: -0.0000

Data Index: 81834 (C4 (Web Text))

Max Activating Token Index: 150

Text #13

Max Range: 1.1249. Min Range: -1.1249

Max Act: 0.5566. Min Act: -0.0000

Data Index: 871519 (C4 (Web Text))

Max Activating Token Index: 316

Text #14

Max Range: 1.1249. Min Range: -1.1249

Max Act: 0.5706. Min Act: -0.0000

Data Index: 722324 (C4 (Web Text))

Max Activating Token Index: 191

Text #15

Max Range: 1.1249. Min Range: -1.1249

Max Act: 0.5368. Min Act: -0.0000

Data Index: 472582 (C4 (Web Text))

Max Activating Token Index: 621

Text #16

Max Range: 1.1249. Min Range: -1.1249

Max Act: 0.5125. Min Act: -0.0000

Data Index: 495512 (C4 (Web Text))

Max Activating Token Index: 216

Text #17

Max Range: 1.1249. Min Range: -1.1249

Max Act: 0.5541. Min Act: -0.0000

Data Index: 1342024 (C4 (Web Text))

Max Activating Token Index: 896

Text #18

Max Range: 1.1249. Min Range: -1.1249

Max Act: 0.5153. Min Act: -0.0001

Data Index: 1512163 (Python Code)

Max Activating Token Index: 351

Text #19

Max Range: 1.1249. Min Range: -1.1249

Max Act: 0.4883. Min Act: -0.0001

Data Index: 1385227 (Python Code)

Max Activating Token Index: 1008