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

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

Neuron 3433 in Layer 2

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 1.5936. Min Range: -1.5936

Max Act: 1.5349. Min Act: -0.0001

Data Index: 85642 (C4 (Web Text))

Max Activating Token Index: 174

Text #1

Max Range: 1.5936. Min Range: -1.5936

Max Act: 1.5316. Min Act: -0.0001

Data Index: 900553 (C4 (Web Text))

Max Activating Token Index: 647

Text #2

Max Range: 1.5936. Min Range: -1.5936

Max Act: 1.4785. Min Act: -0.0001

Data Index: 1160887 (C4 (Web Text))

Max Activating Token Index: 692

Text #3

Max Range: 1.5936. Min Range: -1.5936

Max Act: 1.5936. Min Act: -0.0002

Data Index: 47413 (C4 (Web Text))

Max Activating Token Index: 837

Text #4

Max Range: 1.5936. Min Range: -1.5936

Max Act: 1.3673. Min Act: -0.0002

Data Index: 247343 (C4 (Web Text))

Max Activating Token Index: 39

Text #5

Max Range: 1.5936. Min Range: -1.5936

Max Act: 1.3700. Min Act: -0.0002

Data Index: 195027 (C4 (Web Text))

Max Activating Token Index: 636

Text #6

Max Range: 1.5936. Min Range: -1.5936

Max Act: 1.2494. Min Act: -0.0002

Data Index: 158764 (C4 (Web Text))

Max Activating Token Index: 448

Text #7

Max Range: 1.5936. Min Range: -1.5936

Max Act: 1.2732. Min Act: -0.0002

Data Index: 582648 (C4 (Web Text))

Max Activating Token Index: 473

Text #8

Max Range: 1.5936. Min Range: -1.5936

Max Act: 1.1848. Min Act: -0.0002

Data Index: 825650 (C4 (Web Text))

Max Activating Token Index: 741

Text #9

Max Range: 1.5936. Min Range: -1.5936

Max Act: 1.2204. Min Act: -0.0002

Data Index: 167296 (C4 (Web Text))

Max Activating Token Index: 639

Text #10

Max Range: 1.5936. Min Range: -1.5936

Max Act: 1.1789. Min Act: -0.0001

Data Index: 1133614 (C4 (Web Text))

Max Activating Token Index: 281

Text #11

Max Range: 1.5936. Min Range: -1.5936

Max Act: 1.1415. Min Act: -0.0002

Data Index: 86452 (C4 (Web Text))

Max Activating Token Index: 845

Text #12

Max Range: 1.5936. Min Range: -1.5936

Max Act: 1.2195. Min Act: -0.0002

Data Index: 1041472 (C4 (Web Text))

Max Activating Token Index: 854

Text #13

Max Range: 1.5936. Min Range: -1.5936

Max Act: 1.2380. Min Act: -0.0002

Data Index: 655485 (C4 (Web Text))

Max Activating Token Index: 240

Text #14

Max Range: 1.5936. Min Range: -1.5936

Max Act: 1.1686. Min Act: -0.0002

Data Index: 567931 (C4 (Web Text))

Max Activating Token Index: 370

Text #15

Max Range: 1.5936. Min Range: -1.5936

Max Act: 1.1292. Min Act: -0.0002

Data Index: 260446 (C4 (Web Text))

Max Activating Token Index: 489

Text #16

Max Range: 1.5936. Min Range: -1.5936

Max Act: 1.0857. Min Act: -0.0002

Data Index: 983234 (C4 (Web Text))

Max Activating Token Index: 225

Text #17

Max Range: 1.5936. Min Range: -1.5936

Max Act: 1.0522. Min Act: -0.0002

Data Index: 414906 (C4 (Web Text))

Max Activating Token Index: 579

Text #18

Max Range: 1.5936. Min Range: -1.5936

Max Act: 1.0470. Min Act: -0.0001

Data Index: 1431716 (Python Code)

Max Activating Token Index: 711

Text #19

Max Range: 1.5936. Min Range: -1.5936

Max Act: 1.0437. Min Act: -0.0001

Data Index: 1489800 (Python Code)

Max Activating Token Index: 647