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

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

Neuron 20 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: 0.0548. Min Range: -0.0548

Max Act: 0.0528. Min Act: -0.0001

Data Index: 553955 (C4 (Web Text))

Max Activating Token Index: 687

Text #1

Max Range: 0.0548. Min Range: -0.0548

Max Act: 0.0548. Min Act: -0.0001

Data Index: 1149125 (C4 (Web Text))

Max Activating Token Index: 755

Text #2

Max Range: 0.0548. Min Range: -0.0548

Max Act: 0.0504. Min Act: -0.0002

Data Index: 1158059 (C4 (Web Text))

Max Activating Token Index: 921

Text #3

Max Range: 0.0548. Min Range: -0.0548

Max Act: 0.0487. Min Act: -0.0001

Data Index: 732913 (C4 (Web Text))

Max Activating Token Index: 879

Text #4

Max Range: 0.0548. Min Range: -0.0548

Max Act: 0.0485. Min Act: -0.0002

Data Index: 970283 (C4 (Web Text))

Max Activating Token Index: 293

Text #5

Max Range: 0.0548. Min Range: -0.0548

Max Act: 0.0487. Min Act: -0.0002

Data Index: 475849 (C4 (Web Text))

Max Activating Token Index: 100

Text #6

Max Range: 0.0548. Min Range: -0.0548

Max Act: 0.0494. Min Act: -0.0002

Data Index: 274650 (C4 (Web Text))

Max Activating Token Index: 783

Text #7

Max Range: 0.0548. Min Range: -0.0548

Max Act: 0.0477. Min Act: -0.0002

Data Index: 971653 (C4 (Web Text))

Max Activating Token Index: 301

Text #8

Max Range: 0.0548. Min Range: -0.0548

Max Act: 0.0482. Min Act: -0.0002

Data Index: 29400 (C4 (Web Text))

Max Activating Token Index: 850

Text #9

Max Range: 0.0548. Min Range: -0.0548

Max Act: 0.0478. Min Act: -0.0001

Data Index: 482402 (C4 (Web Text))

Max Activating Token Index: 595

Text #10

Max Range: 0.0548. Min Range: -0.0548

Max Act: 0.0480. Min Act: -0.0001

Data Index: 673725 (C4 (Web Text))

Max Activating Token Index: 226

Text #11

Max Range: 0.0548. Min Range: -0.0548

Max Act: 0.0451. Min Act: -0.0002

Data Index: 974101 (C4 (Web Text))

Max Activating Token Index: 747

Text #12

Max Range: 0.0548. Min Range: -0.0548

Max Act: 0.0466. Min Act: -0.0001

Data Index: 130187 (C4 (Web Text))

Max Activating Token Index: 782

Text #13

Max Range: 0.0548. Min Range: -0.0548

Max Act: 0.0467. Min Act: -0.0002

Data Index: 974058 (C4 (Web Text))

Max Activating Token Index: 255

Text #14

Max Range: 0.0548. Min Range: -0.0548

Max Act: 0.0449. Min Act: -0.0002

Data Index: 509274 (C4 (Web Text))

Max Activating Token Index: 143

Text #15

Max Range: 0.0548. Min Range: -0.0548

Max Act: 0.0453. Min Act: -0.0002

Data Index: 705154 (C4 (Web Text))

Max Activating Token Index: 30

Text #16

Max Range: 0.0548. Min Range: -0.0548

Max Act: 0.0457. Min Act: -0.0002

Data Index: 1018128 (C4 (Web Text))

Max Activating Token Index: 858

Text #17

Max Range: 0.0548. Min Range: -0.0548

Max Act: 0.0466. Min Act: -0.0002

Data Index: 476662 (C4 (Web Text))

Max Activating Token Index: 56

Text #18

Max Range: 0.0548. Min Range: -0.0548

Max Act: 0.0451. Min Act: -0.0002

Data Index: 386043 (C4 (Web Text))

Max Activating Token Index: 844

Text #19

Max Range: 0.0548. Min Range: -0.0548

Max Act: 0.0443. Min Act: -0.0002

Data Index: 424056 (C4 (Web Text))

Max Activating Token Index: 311