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

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

Neuron 4198 in Layer 4

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.6403. Min Range: -0.6403

Max Act: 0.6319. Min Act: -0.0000

Data Index: 1562079 (Python Code)

Max Activating Token Index: 723

Text #1

Max Range: 0.6403. Min Range: -0.6403

Max Act: 0.6403. Min Act: -0.0001

Data Index: 743958 (C4 (Web Text))

Max Activating Token Index: 131

Text #2

Max Range: 0.6403. Min Range: -0.6403

Max Act: 0.6082. Min Act: -0.0000

Data Index: 1595194 (Python Code)

Max Activating Token Index: 831

Text #3

Max Range: 0.6403. Min Range: -0.6403

Max Act: 0.5733. Min Act: -0.0001

Data Index: 1346559 (C4 (Web Text))

Max Activating Token Index: 28

Text #4

Max Range: 0.6403. Min Range: -0.6403

Max Act: 0.6307. Min Act: -0.0001

Data Index: 978132 (C4 (Web Text))

Max Activating Token Index: 139

Text #5

Max Range: 0.6403. Min Range: -0.6403

Max Act: 0.5743. Min Act: -0.0001

Data Index: 595211 (C4 (Web Text))

Max Activating Token Index: 502

Text #6

Max Range: 0.6403. Min Range: -0.6403

Max Act: 0.5656. Min Act: -0.0001

Data Index: 276921 (C4 (Web Text))

Max Activating Token Index: 903

Text #7

Max Range: 0.6403. Min Range: -0.6403

Max Act: 0.5549. Min Act: -0.0001

Data Index: 522972 (C4 (Web Text))

Max Activating Token Index: 696

Text #8

Max Range: 0.6403. Min Range: -0.6403

Max Act: 0.5525. Min Act: -0.0001

Data Index: 1158622 (C4 (Web Text))

Max Activating Token Index: 64

Text #9

Max Range: 0.6403. Min Range: -0.6403

Max Act: 0.5369. Min Act: -0.0001

Data Index: 743351 (C4 (Web Text))

Max Activating Token Index: 210

Text #10

Max Range: 0.6403. Min Range: -0.6403

Max Act: 0.5312. Min Act: -0.0001

Data Index: 36114 (C4 (Web Text))

Max Activating Token Index: 169

Text #11

Max Range: 0.6403. Min Range: -0.6403

Max Act: 0.5189. Min Act: -0.0001

Data Index: 379494 (C4 (Web Text))

Max Activating Token Index: 921

Text #12

Max Range: 0.6403. Min Range: -0.6403

Max Act: 0.5371. Min Act: -0.0001

Data Index: 974912 (C4 (Web Text))

Max Activating Token Index: 889

Text #13

Max Range: 0.6403. Min Range: -0.6403

Max Act: 0.5206. Min Act: -0.0001

Data Index: 411488 (C4 (Web Text))

Max Activating Token Index: 232

Text #14

Max Range: 0.6403. Min Range: -0.6403

Max Act: 0.4964. Min Act: -0.0001

Data Index: 614900 (C4 (Web Text))

Max Activating Token Index: 756

Text #15

Max Range: 0.6403. Min Range: -0.6403

Max Act: 0.5266. Min Act: -0.0001

Data Index: 1234839 (C4 (Web Text))

Max Activating Token Index: 884

Text #16

Max Range: 0.6403. Min Range: -0.6403

Max Act: 0.5099. Min Act: -0.0001

Data Index: 40266 (C4 (Web Text))

Max Activating Token Index: 362

Text #17

Max Range: 0.6403. Min Range: -0.6403

Max Act: 0.5050. Min Act: -0.0000

Data Index: 1520655 (Python Code)

Max Activating Token Index: 863

Text #18

Max Range: 0.6403. Min Range: -0.6403

Max Act: 0.5186. Min Act: -0.0001

Data Index: 281025 (C4 (Web Text))

Max Activating Token Index: 465

Text #19

Max Range: 0.6403. Min Range: -0.6403

Max Act: 0.4966. Min Act: -0.0001

Data Index: 808884 (C4 (Web Text))

Max Activating Token Index: 437