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

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

Neuron 3029 in Layer 3

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.5420. Min Range: -0.5420

Max Act: 0.5420. Min Act: -0.0001

Data Index: 1128532 (C4 (Web Text))

Max Activating Token Index: 2

Text #1

Max Range: 0.5420. Min Range: -0.5420

Max Act: 0.4904. Min Act: -0.0001

Data Index: 1548183 (Python Code)

Max Activating Token Index: 806

Text #2

Max Range: 0.5420. Min Range: -0.5420

Max Act: 0.4688. Min Act: -0.0001

Data Index: 303997 (C4 (Web Text))

Max Activating Token Index: 934

Text #3

Max Range: 0.5420. Min Range: -0.5420

Max Act: 0.4897. Min Act: -0.0001

Data Index: 1244970 (C4 (Web Text))

Max Activating Token Index: 52

Text #4

Max Range: 0.5420. Min Range: -0.5420

Max Act: 0.4999. Min Act: -0.0001

Data Index: 1316387 (C4 (Web Text))

Max Activating Token Index: 25

Text #5

Max Range: 0.5420. Min Range: -0.5420

Max Act: 0.4868. Min Act: -0.0001

Data Index: 476114 (C4 (Web Text))

Max Activating Token Index: 672

Text #6

Max Range: 0.5420. Min Range: -0.5420

Max Act: 0.4550. Min Act: -0.0001

Data Index: 1196405 (C4 (Web Text))

Max Activating Token Index: 239

Text #7

Max Range: 0.5420. Min Range: -0.5420

Max Act: 0.4560. Min Act: -0.0001

Data Index: 1394748 (Python Code)

Max Activating Token Index: 844

Text #8

Max Range: 0.5420. Min Range: -0.5420

Max Act: 0.4657. Min Act: -0.0001

Data Index: 427757 (C4 (Web Text))

Max Activating Token Index: 13

Text #9

Max Range: 0.5420. Min Range: -0.5420

Max Act: 0.4546. Min Act: -0.0001

Data Index: 543857 (C4 (Web Text))

Max Activating Token Index: 26

Text #10

Max Range: 0.5420. Min Range: -0.5420

Max Act: 0.5124. Min Act: -0.0001

Data Index: 1488963 (Python Code)

Max Activating Token Index: 220

Text #11

Max Range: 0.5420. Min Range: -0.5420

Max Act: 0.4663. Min Act: -0.0001

Data Index: 1504157 (Python Code)

Max Activating Token Index: 40

Text #12

Max Range: 0.5420. Min Range: -0.5420

Max Act: 0.3820. Min Act: -0.0001

Data Index: 1491416 (Python Code)

Max Activating Token Index: 733

Text #13

Max Range: 0.5420. Min Range: -0.5420

Max Act: 0.4183. Min Act: -0.0001

Data Index: 1408676 (Python Code)

Max Activating Token Index: 36

Text #14

Max Range: 0.5420. Min Range: -0.5420

Max Act: 0.4342. Min Act: -0.0001

Data Index: 138844 (C4 (Web Text))

Max Activating Token Index: 348

Text #15

Max Range: 0.5420. Min Range: -0.5420

Max Act: 0.4410. Min Act: -0.0001

Data Index: 1494594 (Python Code)

Max Activating Token Index: 880

Text #16

Max Range: 0.5420. Min Range: -0.5420

Max Act: 0.4271. Min Act: -0.0001

Data Index: 833245 (C4 (Web Text))

Max Activating Token Index: 696

Text #17

Max Range: 0.5420. Min Range: -0.5420

Max Act: 0.3861. Min Act: -0.0001

Data Index: 775090 (C4 (Web Text))

Max Activating Token Index: 823

Text #18

Max Range: 0.5420. Min Range: -0.5420

Max Act: 0.4360. Min Act: -0.0001

Data Index: 70275 (C4 (Web Text))

Max Activating Token Index: 41

Text #19

Max Range: 0.5420. Min Range: -0.5420

Max Act: 0.4151. Min Act: -0.0001

Data Index: 73287 (C4 (Web Text))

Max Activating Token Index: 34