Model: SoLU Model: 8 Layers, 4096 Neurons per Layer

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

Neuron 3537 in Layer 6

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-8l')

Text #0

Max Range: 0.6612. Min Range: -0.6612

Max Act: 0.6612. Min Act: 0.0000

Data Index: 1631747 (Python Code)

Max Activating Token Index: 151

Text #1

Max Range: 0.6612. Min Range: -0.6612

Max Act: 0.6282. Min Act: -0.0000

Data Index: 891245 (C4 (Web Text))

Max Activating Token Index: 674

Text #2

Max Range: 0.6612. Min Range: -0.6612

Max Act: 0.6426. Min Act: -0.0001

Data Index: 540588 (C4 (Web Text))

Max Activating Token Index: 443

Text #3

Max Range: 0.6612. Min Range: -0.6612

Max Act: 0.5518. Min Act: -0.0001

Data Index: 891152 (C4 (Web Text))

Max Activating Token Index: 898

Text #4

Max Range: 0.6612. Min Range: -0.6612

Max Act: 0.5337. Min Act: -0.0001

Data Index: 64979 (C4 (Web Text))

Max Activating Token Index: 103

Text #5

Max Range: 0.6612. Min Range: -0.6612

Max Act: 0.5367. Min Act: -0.0001

Data Index: 488164 (C4 (Web Text))

Max Activating Token Index: 250

Text #6

Max Range: 0.6612. Min Range: -0.6612

Max Act: 0.5521. Min Act: -0.0001

Data Index: 1052599 (C4 (Web Text))

Max Activating Token Index: 694

Text #7

Max Range: 0.6612. Min Range: -0.6612

Max Act: 0.5637. Min Act: -0.0000

Data Index: 1391823 (Python Code)

Max Activating Token Index: 136

Text #8

Max Range: 0.6612. Min Range: -0.6612

Max Act: 0.5015. Min Act: -0.0000

Data Index: 910467 (C4 (Web Text))

Max Activating Token Index: 849

Text #9

Max Range: 0.6612. Min Range: -0.6612

Max Act: 0.5794. Min Act: -0.0001

Data Index: 281200 (C4 (Web Text))

Max Activating Token Index: 878

Text #10

Max Range: 0.6612. Min Range: -0.6612

Max Act: 0.5006. Min Act: -0.0000

Data Index: 939450 (C4 (Web Text))

Max Activating Token Index: 618

Text #11

Max Range: 0.6612. Min Range: -0.6612

Max Act: 0.4791. Min Act: -0.0000

Data Index: 1406178 (Python Code)

Max Activating Token Index: 487

Text #12

Max Range: 0.6612. Min Range: -0.6612

Max Act: 0.4890. Min Act: -0.0001

Data Index: 327369 (C4 (Web Text))

Max Activating Token Index: 1002

Text #13

Max Range: 0.6612. Min Range: -0.6612

Max Act: 0.5127. Min Act: -0.0000

Data Index: 1316229 (C4 (Web Text))

Max Activating Token Index: 719

Text #14

Max Range: 0.6612. Min Range: -0.6612

Max Act: 0.4659. Min Act: -0.0001

Data Index: 417489 (C4 (Web Text))

Max Activating Token Index: 920

Text #15

Max Range: 0.6612. Min Range: -0.6612

Max Act: 0.4779. Min Act: -0.0001

Data Index: 49714 (C4 (Web Text))

Max Activating Token Index: 627

Text #16

Max Range: 0.6612. Min Range: -0.6612

Max Act: 0.4808. Min Act: -0.0001

Data Index: 638764 (C4 (Web Text))

Max Activating Token Index: 823

Text #17

Max Range: 0.6612. Min Range: -0.6612

Max Act: 0.4761. Min Act: -0.0000

Data Index: 476335 (C4 (Web Text))

Max Activating Token Index: 337

Text #18

Max Range: 0.6612. Min Range: -0.6612

Max Act: 0.5202. Min Act: -0.0000

Data Index: 731134 (C4 (Web Text))

Max Activating Token Index: 55

Text #19

Max Range: 0.6612. Min Range: -0.6612

Max Act: 0.5009. Min Act: -0.0001

Data Index: 834673 (C4 (Web Text))

Max Activating Token Index: 801