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

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

Neuron 4240 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: 5.8298. Min Range: -5.8298

Max Act: 5.8298. Min Act: -0.0000

Data Index: 805170 (C4 (Web Text))

Max Activating Token Index: 255

Text #1

Max Range: 5.8298. Min Range: -5.8298

Max Act: 5.8169. Min Act: -0.0000

Data Index: 802819 (C4 (Web Text))

Max Activating Token Index: 330

Text #2

Max Range: 5.8298. Min Range: -5.8298

Max Act: 5.3896. Min Act: -0.0000

Data Index: 644142 (C4 (Web Text))

Max Activating Token Index: 217

Text #3

Max Range: 5.8298. Min Range: -5.8298

Max Act: 5.2009. Min Act: -0.0000

Data Index: 889672 (C4 (Web Text))

Max Activating Token Index: 941

Text #4

Max Range: 5.8298. Min Range: -5.8298

Max Act: 5.1502. Min Act: -0.0000

Data Index: 32052 (C4 (Web Text))

Max Activating Token Index: 125

Text #5

Max Range: 5.8298. Min Range: -5.8298

Max Act: 4.7379. Min Act: -0.0000

Data Index: 797284 (C4 (Web Text))

Max Activating Token Index: 727

Text #6

Max Range: 5.8298. Min Range: -5.8298

Max Act: 4.8836. Min Act: -0.0000

Data Index: 1122151 (C4 (Web Text))

Max Activating Token Index: 93

Text #7

Max Range: 5.8298. Min Range: -5.8298

Max Act: 4.2389. Min Act: -0.0000

Data Index: 804423 (C4 (Web Text))

Max Activating Token Index: 643

Text #8

Max Range: 5.8298. Min Range: -5.8298

Max Act: 4.0542. Min Act: -0.0000

Data Index: 1242582 (C4 (Web Text))

Max Activating Token Index: 524

Text #9

Max Range: 5.8298. Min Range: -5.8298

Max Act: 4.4898. Min Act: -0.0000

Data Index: 1079079 (C4 (Web Text))

Max Activating Token Index: 132

Text #10

Max Range: 5.8298. Min Range: -5.8298

Max Act: 4.3099. Min Act: -0.0000

Data Index: 393312 (C4 (Web Text))

Max Activating Token Index: 754

Text #11

Max Range: 5.8298. Min Range: -5.8298

Max Act: 4.0702. Min Act: -0.0000

Data Index: 1616738 (Python Code)

Max Activating Token Index: 928

Text #12

Max Range: 5.8298. Min Range: -5.8298

Max Act: 4.0410. Min Act: -0.0000

Data Index: 815510 (C4 (Web Text))

Max Activating Token Index: 155

Text #13

Max Range: 5.8298. Min Range: -5.8298

Max Act: 3.8136. Min Act: -0.0000

Data Index: 127758 (C4 (Web Text))

Max Activating Token Index: 207

Text #14

Max Range: 5.8298. Min Range: -5.8298

Max Act: 3.6509. Min Act: -0.0000

Data Index: 1417298 (Python Code)

Max Activating Token Index: 307

Text #15

Max Range: 5.8298. Min Range: -5.8298

Max Act: 3.4683. Min Act: -0.0000

Data Index: 497067 (C4 (Web Text))

Max Activating Token Index: 482

Text #16

Max Range: 5.8298. Min Range: -5.8298

Max Act: 3.5977. Min Act: -0.0000

Data Index: 702565 (C4 (Web Text))

Max Activating Token Index: 533

Text #17

Max Range: 5.8298. Min Range: -5.8298

Max Act: 3.5307. Min Act: -0.0000

Data Index: 1370421 (Python Code)

Max Activating Token Index: 146

Text #18

Max Range: 5.8298. Min Range: -5.8298

Max Act: 3.5399. Min Act: -0.0000

Data Index: 1302972 (C4 (Web Text))

Max Activating Token Index: 317

Text #19

Max Range: 5.8298. Min Range: -5.8298

Max Act: 3.4672. Min Act: -0.0000

Data Index: 1466247 (Python Code)

Max Activating Token Index: 920