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

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

Neuron 1058 in Layer 7

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 5.1140. Min Range: -5.1140

Max Act: 5.1140. Min Act: -0.0000

Data Index: 1485587 (Python Code)

Max Activating Token Index: 978

Text #1

Max Range: 5.1140. Min Range: -5.1140

Max Act: 4.6184. Min Act: -0.0000

Data Index: 1460112 (Python Code)

Max Activating Token Index: 1001

Text #2

Max Range: 5.1140. Min Range: -5.1140

Max Act: 4.5812. Min Act: -0.0000

Data Index: 190347 (C4 (Web Text))

Max Activating Token Index: 750

Text #3

Max Range: 5.1140. Min Range: -5.1140

Max Act: 4.4489. Min Act: -0.0000

Data Index: 43652 (C4 (Web Text))

Max Activating Token Index: 160

Text #4

Max Range: 5.1140. Min Range: -5.1140

Max Act: 4.1599. Min Act: -0.0000

Data Index: 1602473 (Python Code)

Max Activating Token Index: 767

Text #5

Max Range: 5.1140. Min Range: -5.1140

Max Act: 4.1516. Min Act: -0.0000

Data Index: 1309998 (C4 (Web Text))

Max Activating Token Index: 232

Text #6

Max Range: 5.1140. Min Range: -5.1140

Max Act: 4.1391. Min Act: -0.0000

Data Index: 1535150 (Python Code)

Max Activating Token Index: 647

Text #7

Max Range: 5.1140. Min Range: -5.1140

Max Act: 4.1274. Min Act: -0.0000

Data Index: 1409227 (Python Code)

Max Activating Token Index: 863

Text #8

Max Range: 5.1140. Min Range: -5.1140

Max Act: 4.1034. Min Act: -0.0000

Data Index: 1566237 (Python Code)

Max Activating Token Index: 1008

Text #9

Max Range: 5.1140. Min Range: -5.1140

Max Act: 3.7744. Min Act: -0.0000

Data Index: 1575423 (Python Code)

Max Activating Token Index: 192

Text #10

Max Range: 5.1140. Min Range: -5.1140

Max Act: 4.1000. Min Act: -0.0000

Data Index: 527831 (C4 (Web Text))

Max Activating Token Index: 1001

Text #11

Max Range: 5.1140. Min Range: -5.1140

Max Act: 4.1311. Min Act: -0.0000

Data Index: 230489 (C4 (Web Text))

Max Activating Token Index: 144

Text #12

Max Range: 5.1140. Min Range: -5.1140

Max Act: 3.7374. Min Act: -0.0000

Data Index: 1530125 (Python Code)

Max Activating Token Index: 208

Text #13

Max Range: 5.1140. Min Range: -5.1140

Max Act: 3.7194. Min Act: -0.0000

Data Index: 1644669 (Python Code)

Max Activating Token Index: 528

Text #14

Max Range: 5.1140. Min Range: -5.1140

Max Act: 4.0890. Min Act: -0.0000

Data Index: 1502837 (Python Code)

Max Activating Token Index: 604

Text #15

Max Range: 5.1140. Min Range: -5.1140

Max Act: 3.8893. Min Act: -0.0000

Data Index: 744680 (C4 (Web Text))

Max Activating Token Index: 802

Text #16

Max Range: 5.1140. Min Range: -5.1140

Max Act: 4.0104. Min Act: -0.0000

Data Index: 870686 (C4 (Web Text))

Max Activating Token Index: 557

Text #17

Max Range: 5.1140. Min Range: -5.1140

Max Act: 3.8179. Min Act: -0.0000

Data Index: 1508265 (Python Code)

Max Activating Token Index: 253

Text #18

Max Range: 5.1140. Min Range: -5.1140

Max Act: 3.7587. Min Act: -0.0000

Data Index: 1225710 (C4 (Web Text))

Max Activating Token Index: 620

Text #19

Max Range: 5.1140. Min Range: -5.1140

Max Act: 3.3898. Min Act: -0.0000

Data Index: 1599597 (Python Code)

Max Activating Token Index: 348