Model: SoLU Model: 4 Layers, 2048 Neurons per Layer

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

Neuron 89 in Layer 3

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-4l')

Text #0

Max Range: 0.0508. Min Range: -0.0508

Max Act: 0.0508. Min Act: -0.0001

Data Index: 459384 (C4 (Web Text))

Max Activating Token Index: 88

Text #1

Max Range: 0.0508. Min Range: -0.0508

Max Act: 0.0506. Min Act: -0.0002

Data Index: 961783 (C4 (Web Text))

Max Activating Token Index: 810

Text #2

Max Range: 0.0508. Min Range: -0.0508

Max Act: 0.0501. Min Act: -0.0001

Data Index: 369766 (C4 (Web Text))

Max Activating Token Index: 649

Text #3

Max Range: 0.0508. Min Range: -0.0508

Max Act: 0.0500. Min Act: -0.0002

Data Index: 380335 (C4 (Web Text))

Max Activating Token Index: 605

Text #4

Max Range: 0.0508. Min Range: -0.0508

Max Act: 0.0496. Min Act: -0.0002

Data Index: 1124283 (C4 (Web Text))

Max Activating Token Index: 688

Text #5

Max Range: 0.0508. Min Range: -0.0508

Max Act: 0.0460. Min Act: -0.0002

Data Index: 954739 (C4 (Web Text))

Max Activating Token Index: 628

Text #6

Max Range: 0.0508. Min Range: -0.0508

Max Act: 0.0486. Min Act: -0.0002

Data Index: 114095 (C4 (Web Text))

Max Activating Token Index: 268

Text #7

Max Range: 0.0508. Min Range: -0.0508

Max Act: 0.0506. Min Act: -0.0002

Data Index: 237682 (C4 (Web Text))

Max Activating Token Index: 655

Text #8

Max Range: 0.0508. Min Range: -0.0508

Max Act: 0.0458. Min Act: -0.0001

Data Index: 690913 (C4 (Web Text))

Max Activating Token Index: 366

Text #9

Max Range: 0.0508. Min Range: -0.0508

Max Act: 0.0428. Min Act: -0.0002

Data Index: 640124 (C4 (Web Text))

Max Activating Token Index: 1016

Text #10

Max Range: 0.0508. Min Range: -0.0508

Max Act: 0.0439. Min Act: -0.0002

Data Index: 1334206 (C4 (Web Text))

Max Activating Token Index: 759

Text #11

Max Range: 0.0508. Min Range: -0.0508

Max Act: 0.0446. Min Act: -0.0002

Data Index: 1414441 (Python Code)

Max Activating Token Index: 997

Text #12

Max Range: 0.0508. Min Range: -0.0508

Max Act: 0.0436. Min Act: -0.0002

Data Index: 627216 (C4 (Web Text))

Max Activating Token Index: 879

Text #13

Max Range: 0.0508. Min Range: -0.0508

Max Act: 0.0456. Min Act: -0.0002

Data Index: 1630714 (Python Code)

Max Activating Token Index: 749

Text #14

Max Range: 0.0508. Min Range: -0.0508

Max Act: 0.0422. Min Act: -0.0002

Data Index: 1173168 (C4 (Web Text))

Max Activating Token Index: 360

Text #15

Max Range: 0.0508. Min Range: -0.0508

Max Act: 0.0433. Min Act: -0.0002

Data Index: 473103 (C4 (Web Text))

Max Activating Token Index: 920

Text #16

Max Range: 0.0508. Min Range: -0.0508

Max Act: 0.0425. Min Act: -0.0002

Data Index: 125534 (C4 (Web Text))

Max Activating Token Index: 916

Text #17

Max Range: 0.0508. Min Range: -0.0508

Max Act: 0.0398. Min Act: -0.0002

Data Index: 201187 (C4 (Web Text))

Max Activating Token Index: 554

Text #18

Max Range: 0.0508. Min Range: -0.0508

Max Act: 0.0406. Min Act: -0.0002

Data Index: 721919 (C4 (Web Text))

Max Activating Token Index: 554

Text #19

Max Range: 0.0508. Min Range: -0.0508

Max Act: 0.0402. Min Act: -0.0001

Data Index: 1080652 (C4 (Web Text))

Max Activating Token Index: 419