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

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

Neuron 4631 in Layer 6

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 3.9168. Min Range: -3.9168

Max Act: 3.9168. Min Act: -0.0000

Data Index: 1610755 (Python Code)

Max Activating Token Index: 283

Text #1

Max Range: 3.9168. Min Range: -3.9168

Max Act: 3.5827. Min Act: -0.0000

Data Index: 63228 (C4 (Web Text))

Max Activating Token Index: 768

Text #2

Max Range: 3.9168. Min Range: -3.9168

Max Act: 3.7700. Min Act: -0.0000

Data Index: 1555450 (Python Code)

Max Activating Token Index: 989

Text #3

Max Range: 3.9168. Min Range: -3.9168

Max Act: 3.4579. Min Act: -0.0000

Data Index: 1586339 (Python Code)

Max Activating Token Index: 230

Text #4

Max Range: 3.9168. Min Range: -3.9168

Max Act: 3.2472. Min Act: -0.0000

Data Index: 844044 (C4 (Web Text))

Max Activating Token Index: 461

Text #5

Max Range: 3.9168. Min Range: -3.9168

Max Act: 2.6795. Min Act: -0.0000

Data Index: 1614262 (Python Code)

Max Activating Token Index: 339

Text #6

Max Range: 3.9168. Min Range: -3.9168

Max Act: 3.5148. Min Act: -0.0000

Data Index: 1428860 (Python Code)

Max Activating Token Index: 555

Text #7

Max Range: 3.9168. Min Range: -3.9168

Max Act: 3.3934. Min Act: -0.0000

Data Index: 1451384 (Python Code)

Max Activating Token Index: 165

Text #8

Max Range: 3.9168. Min Range: -3.9168

Max Act: 2.9018. Min Act: -0.0000

Data Index: 1443244 (Python Code)

Max Activating Token Index: 141

Text #9

Max Range: 3.9168. Min Range: -3.9168

Max Act: 2.4598. Min Act: -0.0000

Data Index: 1044238 (C4 (Web Text))

Max Activating Token Index: 351

Text #10

Max Range: 3.9168. Min Range: -3.9168

Max Act: 3.2520. Min Act: -0.0000

Data Index: 1427869 (Python Code)

Max Activating Token Index: 158

Text #11

Max Range: 3.9168. Min Range: -3.9168

Max Act: 2.9451. Min Act: -0.0000

Data Index: 509493 (C4 (Web Text))

Max Activating Token Index: 600

Text #12

Max Range: 3.9168. Min Range: -3.9168

Max Act: 2.9389. Min Act: -0.0000

Data Index: 1606090 (Python Code)

Max Activating Token Index: 517

Text #13

Max Range: 3.9168. Min Range: -3.9168

Max Act: 2.7643. Min Act: -0.0000

Data Index: 198098 (C4 (Web Text))

Max Activating Token Index: 314

Text #14

Max Range: 3.9168. Min Range: -3.9168

Max Act: 2.7273. Min Act: -0.0000

Data Index: 1507704 (Python Code)

Max Activating Token Index: 958

Text #15

Max Range: 3.9168. Min Range: -3.9168

Max Act: 2.6008. Min Act: -0.0000

Data Index: 1394783 (Python Code)

Max Activating Token Index: 271

Text #16

Max Range: 3.9168. Min Range: -3.9168

Max Act: 2.7085. Min Act: -0.0000

Data Index: 524136 (C4 (Web Text))

Max Activating Token Index: 195

Text #17

Max Range: 3.9168. Min Range: -3.9168

Max Act: 2.6427. Min Act: -0.0000

Data Index: 1552208 (Python Code)

Max Activating Token Index: 375

Text #18

Max Range: 3.9168. Min Range: -3.9168

Max Act: 2.7098. Min Act: -0.0000

Data Index: 295170 (C4 (Web Text))

Max Activating Token Index: 558

Text #19

Max Range: 3.9168. Min Range: -3.9168

Max Act: 2.7043. Min Act: -0.0000

Data Index: 716685 (C4 (Web Text))

Max Activating Token Index: 338