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

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

Neuron 1597 in Layer 4

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 4.0914. Min Range: -4.0914

Max Act: 3.9858. Min Act: -0.0001

Data Index: 1088761 (C4 (Web Text))

Max Activating Token Index: 358

Text #1

Max Range: 4.0914. Min Range: -4.0914

Max Act: 3.7774. Min Act: -0.0001

Data Index: 404536 (C4 (Web Text))

Max Activating Token Index: 108

Text #2

Max Range: 4.0914. Min Range: -4.0914

Max Act: 4.0914. Min Act: -0.0001

Data Index: 641986 (C4 (Web Text))

Max Activating Token Index: 640

Text #3

Max Range: 4.0914. Min Range: -4.0914

Max Act: 3.4318. Min Act: -0.0001

Data Index: 106771 (C4 (Web Text))

Max Activating Token Index: 912

Text #4

Max Range: 4.0914. Min Range: -4.0914

Max Act: 3.9105. Min Act: -0.0001

Data Index: 295750 (C4 (Web Text))

Max Activating Token Index: 411

Text #5

Max Range: 4.0914. Min Range: -4.0914

Max Act: 3.7540. Min Act: -0.0001

Data Index: 61408 (C4 (Web Text))

Max Activating Token Index: 850

Text #6

Max Range: 4.0914. Min Range: -4.0914

Max Act: 3.7070. Min Act: -0.0001

Data Index: 1140572 (C4 (Web Text))

Max Activating Token Index: 102

Text #7

Max Range: 4.0914. Min Range: -4.0914

Max Act: 3.5860. Min Act: -0.0001

Data Index: 186096 (C4 (Web Text))

Max Activating Token Index: 242

Text #8

Max Range: 4.0914. Min Range: -4.0914

Max Act: 3.5327. Min Act: -0.0001

Data Index: 274507 (C4 (Web Text))

Max Activating Token Index: 938

Text #9

Max Range: 4.0914. Min Range: -4.0914

Max Act: 3.4843. Min Act: -0.0001

Data Index: 1299099 (C4 (Web Text))

Max Activating Token Index: 565

Text #10

Max Range: 4.0914. Min Range: -4.0914

Max Act: 3.4800. Min Act: -0.0001

Data Index: 371436 (C4 (Web Text))

Max Activating Token Index: 574

Text #11

Max Range: 4.0914. Min Range: -4.0914

Max Act: 3.3012. Min Act: -0.0001

Data Index: 579653 (C4 (Web Text))

Max Activating Token Index: 995

Text #12

Max Range: 4.0914. Min Range: -4.0914

Max Act: 3.1994. Min Act: -0.0001

Data Index: 940518 (C4 (Web Text))

Max Activating Token Index: 443

Text #13

Max Range: 4.0914. Min Range: -4.0914

Max Act: 3.3512. Min Act: -0.0001

Data Index: 1237268 (C4 (Web Text))

Max Activating Token Index: 819

Text #14

Max Range: 4.0914. Min Range: -4.0914

Max Act: 3.6368. Min Act: -0.0001

Data Index: 57710 (C4 (Web Text))

Max Activating Token Index: 905

Text #15

Max Range: 4.0914. Min Range: -4.0914

Max Act: 2.9998. Min Act: -0.0001

Data Index: 115278 (C4 (Web Text))

Max Activating Token Index: 235

Text #16

Max Range: 4.0914. Min Range: -4.0914

Max Act: 3.1626. Min Act: -0.0001

Data Index: 756107 (C4 (Web Text))

Max Activating Token Index: 867

Text #17

Max Range: 4.0914. Min Range: -4.0914

Max Act: 2.8915. Min Act: -0.0001

Data Index: 726324 (C4 (Web Text))

Max Activating Token Index: 431

Text #18

Max Range: 4.0914. Min Range: -4.0914

Max Act: 2.8677. Min Act: -0.0001

Data Index: 280364 (C4 (Web Text))

Max Activating Token Index: 505

Text #19

Max Range: 4.0914. Min Range: -4.0914

Max Act: 3.5277. Min Act: -0.0001

Data Index: 760501 (C4 (Web Text))

Max Activating Token Index: 388