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

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

Neuron 3764 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: 3.4666. Min Range: -3.4666

Max Act: 3.4093. Min Act: -0.0001

Data Index: 1139143 (C4 (Web Text))

Max Activating Token Index: 49

Text #1

Max Range: 3.4666. Min Range: -3.4666

Max Act: 3.2362. Min Act: -0.0001

Data Index: 992689 (C4 (Web Text))

Max Activating Token Index: 555

Text #2

Max Range: 3.4666. Min Range: -3.4666

Max Act: 3.4666. Min Act: -0.0001

Data Index: 125588 (C4 (Web Text))

Max Activating Token Index: 154

Text #3

Max Range: 3.4666. Min Range: -3.4666

Max Act: 3.2712. Min Act: -0.0001

Data Index: 935669 (C4 (Web Text))

Max Activating Token Index: 603

Text #4

Max Range: 3.4666. Min Range: -3.4666

Max Act: 3.2484. Min Act: -0.0001

Data Index: 669357 (C4 (Web Text))

Max Activating Token Index: 1016

Text #5

Max Range: 3.4666. Min Range: -3.4666

Max Act: 3.3328. Min Act: -0.0001

Data Index: 864495 (C4 (Web Text))

Max Activating Token Index: 267

Text #6

Max Range: 3.4666. Min Range: -3.4666

Max Act: 3.1754. Min Act: -0.0001

Data Index: 982571 (C4 (Web Text))

Max Activating Token Index: 840

Text #7

Max Range: 3.4666. Min Range: -3.4666

Max Act: 3.1479. Min Act: -0.0001

Data Index: 1084588 (C4 (Web Text))

Max Activating Token Index: 54

Text #8

Max Range: 3.4666. Min Range: -3.4666

Max Act: 3.1301. Min Act: -0.0001

Data Index: 286900 (C4 (Web Text))

Max Activating Token Index: 32

Text #9

Max Range: 3.4666. Min Range: -3.4666

Max Act: 3.2376. Min Act: -0.0001

Data Index: 778563 (C4 (Web Text))

Max Activating Token Index: 67

Text #10

Max Range: 3.4666. Min Range: -3.4666

Max Act: 2.7687. Min Act: -0.0001

Data Index: 319153 (C4 (Web Text))

Max Activating Token Index: 822

Text #11

Max Range: 3.4666. Min Range: -3.4666

Max Act: 2.9907. Min Act: -0.0001

Data Index: 527749 (C4 (Web Text))

Max Activating Token Index: 285

Text #12

Max Range: 3.4666. Min Range: -3.4666

Max Act: 3.2401. Min Act: -0.0001

Data Index: 586335 (C4 (Web Text))

Max Activating Token Index: 335

Text #13

Max Range: 3.4666. Min Range: -3.4666

Max Act: 2.8056. Min Act: -0.0001

Data Index: 762850 (C4 (Web Text))

Max Activating Token Index: 829

Text #14

Max Range: 3.4666. Min Range: -3.4666

Max Act: 2.8065. Min Act: -0.0001

Data Index: 1317723 (C4 (Web Text))

Max Activating Token Index: 136

Text #15

Max Range: 3.4666. Min Range: -3.4666

Max Act: 2.5595. Min Act: -0.0001

Data Index: 190472 (C4 (Web Text))

Max Activating Token Index: 752

Text #16

Max Range: 3.4666. Min Range: -3.4666

Max Act: 3.0371. Min Act: -0.0001

Data Index: 1236223 (C4 (Web Text))

Max Activating Token Index: 116

Text #17

Max Range: 3.4666. Min Range: -3.4666

Max Act: 2.8276. Min Act: -0.0001

Data Index: 833935 (C4 (Web Text))

Max Activating Token Index: 405

Text #18

Max Range: 3.4666. Min Range: -3.4666

Max Act: 2.6981. Min Act: -0.0001

Data Index: 601963 (C4 (Web Text))

Max Activating Token Index: 462

Text #19

Max Range: 3.4666. Min Range: -3.4666

Max Act: 2.6749. Min Act: -0.0001

Data Index: 424321 (C4 (Web Text))

Max Activating Token Index: 251