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

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

Neuron 2006 in Layer 5

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.8306. Min Range: -0.8306

Max Act: 0.7920. Min Act: -0.0001

Data Index: 24665 (C4 (Web Text))

Max Activating Token Index: 469

Text #1

Max Range: 0.8306. Min Range: -0.8306

Max Act: 0.8147. Min Act: -0.0001

Data Index: 825263 (C4 (Web Text))

Max Activating Token Index: 440

Text #2

Max Range: 0.8306. Min Range: -0.8306

Max Act: 0.8135. Min Act: -0.0001

Data Index: 203120 (C4 (Web Text))

Max Activating Token Index: 670

Text #3

Max Range: 0.8306. Min Range: -0.8306

Max Act: 0.7697. Min Act: -0.0001

Data Index: 477932 (C4 (Web Text))

Max Activating Token Index: 673

Text #4

Max Range: 0.8306. Min Range: -0.8306

Max Act: 0.7821. Min Act: -0.0001

Data Index: 313350 (C4 (Web Text))

Max Activating Token Index: 735

Text #5

Max Range: 0.8306. Min Range: -0.8306

Max Act: 0.8306. Min Act: -0.0001

Data Index: 825014 (C4 (Web Text))

Max Activating Token Index: 1012

Text #6

Max Range: 0.8306. Min Range: -0.8306

Max Act: 0.8286. Min Act: -0.0001

Data Index: 1015662 (C4 (Web Text))

Max Activating Token Index: 713

Text #7

Max Range: 0.8306. Min Range: -0.8306

Max Act: 0.8010. Min Act: -0.0001

Data Index: 153690 (C4 (Web Text))

Max Activating Token Index: 736

Text #8

Max Range: 0.8306. Min Range: -0.8306

Max Act: 0.7890. Min Act: -0.0001

Data Index: 743242 (C4 (Web Text))

Max Activating Token Index: 149

Text #9

Max Range: 0.8306. Min Range: -0.8306

Max Act: 0.7798. Min Act: -0.0001

Data Index: 716721 (C4 (Web Text))

Max Activating Token Index: 756

Text #10

Max Range: 0.8306. Min Range: -0.8306

Max Act: 0.8004. Min Act: -0.0001

Data Index: 292873 (C4 (Web Text))

Max Activating Token Index: 254

Text #11

Max Range: 0.8306. Min Range: -0.8306

Max Act: 0.7915. Min Act: -0.0001

Data Index: 569933 (C4 (Web Text))

Max Activating Token Index: 891

Text #12

Max Range: 0.8306. Min Range: -0.8306

Max Act: 0.7619. Min Act: -0.0001

Data Index: 891152 (C4 (Web Text))

Max Activating Token Index: 833

Text #13

Max Range: 0.8306. Min Range: -0.8306

Max Act: 0.7760. Min Act: -0.0001

Data Index: 926644 (C4 (Web Text))

Max Activating Token Index: 63

Text #14

Max Range: 0.8306. Min Range: -0.8306

Max Act: 0.7612. Min Act: -0.0001

Data Index: 1589217 (Python Code)

Max Activating Token Index: 175

Text #15

Max Range: 0.8306. Min Range: -0.8306

Max Act: 0.7576. Min Act: -0.0001

Data Index: 722250 (C4 (Web Text))

Max Activating Token Index: 122

Text #16

Max Range: 0.8306. Min Range: -0.8306

Max Act: 0.7583. Min Act: -0.0001

Data Index: 550976 (C4 (Web Text))

Max Activating Token Index: 469

Text #17

Max Range: 0.8306. Min Range: -0.8306

Max Act: 0.7409. Min Act: -0.0001

Data Index: 794928 (C4 (Web Text))

Max Activating Token Index: 1021

Text #18

Max Range: 0.8306. Min Range: -0.8306

Max Act: 0.7448. Min Act: -0.0000

Data Index: 89627 (C4 (Web Text))

Max Activating Token Index: 263

Text #19

Max Range: 0.8306. Min Range: -0.8306

Max Act: 0.7548. Min Act: -0.0001

Data Index: 139646 (C4 (Web Text))

Max Activating Token Index: 680