Model: SoLU Model: 3 Layers, 2048 Neurons per Layer

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

Neuron 128 in Layer 1

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-3l')

Text #0

Max Range: 0.0504. Min Range: -0.0504

Max Act: 0.0499. Min Act: -0.0006

Data Index: 885463 (C4 (Web Text))

Max Activating Token Index: 417

Text #1

Max Range: 0.0504. Min Range: -0.0504

Max Act: 0.0498. Min Act: -0.0006

Data Index: 316351 (C4 (Web Text))

Max Activating Token Index: 500

Text #2

Max Range: 0.0504. Min Range: -0.0504

Max Act: 0.0504. Min Act: -0.0006

Data Index: 178741 (C4 (Web Text))

Max Activating Token Index: 924

Text #3

Max Range: 0.0504. Min Range: -0.0504

Max Act: 0.0498. Min Act: -0.0006

Data Index: 160606 (C4 (Web Text))

Max Activating Token Index: 384

Text #4

Max Range: 0.0504. Min Range: -0.0504

Max Act: 0.0457. Min Act: -0.0006

Data Index: 1185911 (C4 (Web Text))

Max Activating Token Index: 904

Text #5

Max Range: 0.0504. Min Range: -0.0504

Max Act: 0.0421. Min Act: -0.0006

Data Index: 378363 (C4 (Web Text))

Max Activating Token Index: 740

Text #6

Max Range: 0.0504. Min Range: -0.0504

Max Act: 0.0426. Min Act: -0.0006

Data Index: 327081 (C4 (Web Text))

Max Activating Token Index: 812

Text #7

Max Range: 0.0504. Min Range: -0.0504

Max Act: 0.0435. Min Act: -0.0006

Data Index: 838256 (C4 (Web Text))

Max Activating Token Index: 847

Text #8

Max Range: 0.0504. Min Range: -0.0504

Max Act: 0.0424. Min Act: -0.0006

Data Index: 1144584 (C4 (Web Text))

Max Activating Token Index: 629

Text #9

Max Range: 0.0504. Min Range: -0.0504

Max Act: 0.0425. Min Act: -0.0006

Data Index: 1186990 (C4 (Web Text))

Max Activating Token Index: 1004

Text #10

Max Range: 0.0504. Min Range: -0.0504

Max Act: 0.0415. Min Act: -0.0006

Data Index: 473064 (C4 (Web Text))

Max Activating Token Index: 993

Text #11

Max Range: 0.0504. Min Range: -0.0504

Max Act: 0.0412. Min Act: -0.0006

Data Index: 338680 (C4 (Web Text))

Max Activating Token Index: 960

Text #12

Max Range: 0.0504. Min Range: -0.0504

Max Act: 0.0427. Min Act: -0.0006

Data Index: 818340 (C4 (Web Text))

Max Activating Token Index: 653

Text #13

Max Range: 0.0504. Min Range: -0.0504

Max Act: 0.0419. Min Act: -0.0006

Data Index: 144027 (C4 (Web Text))

Max Activating Token Index: 322

Text #14

Max Range: 0.0504. Min Range: -0.0504

Max Act: 0.0396. Min Act: -0.0006

Data Index: 452880 (C4 (Web Text))

Max Activating Token Index: 323

Text #15

Max Range: 0.0504. Min Range: -0.0504

Max Act: 0.0408. Min Act: -0.0006

Data Index: 880216 (C4 (Web Text))

Max Activating Token Index: 692

Text #16

Max Range: 0.0504. Min Range: -0.0504

Max Act: 0.0391. Min Act: -0.0006

Data Index: 709657 (C4 (Web Text))

Max Activating Token Index: 698

Text #17

Max Range: 0.0504. Min Range: -0.0504

Max Act: 0.0400. Min Act: -0.0006

Data Index: 544600 (C4 (Web Text))

Max Activating Token Index: 710

Text #18

Max Range: 0.0504. Min Range: -0.0504

Max Act: 0.0385. Min Act: -0.0006

Data Index: 320909 (C4 (Web Text))

Max Activating Token Index: 62

Text #19

Max Range: 0.0504. Min Range: -0.0504

Max Act: 0.0381. Min Act: -0.0005

Data Index: 509352 (C4 (Web Text))

Max Activating Token Index: 761