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

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

Neuron 655 in Layer 9

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.9507. Min Range: -0.9507

Max Act: 0.9507. Min Act: -0.0000

Data Index: 894255 (C4 (Web Text))

Max Activating Token Index: 844

Text #1

Max Range: 0.9507. Min Range: -0.9507

Max Act: 0.8311. Min Act: -0.0000

Data Index: 50434 (C4 (Web Text))

Max Activating Token Index: 916

Text #2

Max Range: 0.9507. Min Range: -0.9507

Max Act: 0.7881. Min Act: -0.0000

Data Index: 786075 (C4 (Web Text))

Max Activating Token Index: 321

Text #3

Max Range: 0.9507. Min Range: -0.9507

Max Act: 0.7436. Min Act: -0.0000

Data Index: 1334965 (C4 (Web Text))

Max Activating Token Index: 765

Text #4

Max Range: 0.9507. Min Range: -0.9507

Max Act: 0.6826. Min Act: -0.0000

Data Index: 682484 (C4 (Web Text))

Max Activating Token Index: 347

Text #5

Max Range: 0.9507. Min Range: -0.9507

Max Act: 0.6502. Min Act: -0.0000

Data Index: 1206427 (C4 (Web Text))

Max Activating Token Index: 614

Text #6

Max Range: 0.9507. Min Range: -0.9507

Max Act: 0.6393. Min Act: -0.0000

Data Index: 1524378 (Python Code)

Max Activating Token Index: 482

Text #7

Max Range: 0.9507. Min Range: -0.9507

Max Act: 0.6622. Min Act: -0.0000

Data Index: 241547 (C4 (Web Text))

Max Activating Token Index: 720

Text #8

Max Range: 0.9507. Min Range: -0.9507

Max Act: 0.6412. Min Act: -0.0000

Data Index: 903212 (C4 (Web Text))

Max Activating Token Index: 346

Text #9

Max Range: 0.9507. Min Range: -0.9507

Max Act: 0.5743. Min Act: -0.0000

Data Index: 1399236 (Python Code)

Max Activating Token Index: 159

Text #10

Max Range: 0.9507. Min Range: -0.9507

Max Act: 0.6489. Min Act: -0.0000

Data Index: 808124 (C4 (Web Text))

Max Activating Token Index: 978

Text #11

Max Range: 0.9507. Min Range: -0.9507

Max Act: 0.6410. Min Act: -0.0000

Data Index: 640920 (C4 (Web Text))

Max Activating Token Index: 774

Text #12

Max Range: 0.9507. Min Range: -0.9507

Max Act: 0.6155. Min Act: 0.0000

Data Index: 1000253 (C4 (Web Text))

Max Activating Token Index: 416

Text #13

Max Range: 0.9507. Min Range: -0.9507

Max Act: 0.6094. Min Act: -0.0000

Data Index: 363761 (C4 (Web Text))

Max Activating Token Index: 107

Text #14

Max Range: 0.9507. Min Range: -0.9507

Max Act: 0.6222. Min Act: -0.0000

Data Index: 1268405 (C4 (Web Text))

Max Activating Token Index: 697

Text #15

Max Range: 0.9507. Min Range: -0.9507

Max Act: 0.5374. Min Act: -0.0000

Data Index: 1457524 (Python Code)

Max Activating Token Index: 750

Text #16

Max Range: 0.9507. Min Range: -0.9507

Max Act: 0.5987. Min Act: -0.0000

Data Index: 1289190 (C4 (Web Text))

Max Activating Token Index: 1006

Text #17

Max Range: 0.9507. Min Range: -0.9507

Max Act: 0.6072. Min Act: -0.0000

Data Index: 1342618 (C4 (Web Text))

Max Activating Token Index: 921

Text #18

Max Range: 0.9507. Min Range: -0.9507

Max Act: 0.5837. Min Act: -0.0000

Data Index: 279768 (C4 (Web Text))

Max Activating Token Index: 669

Text #19

Max Range: 0.9507. Min Range: -0.9507

Max Act: 0.5991. Min Act: -0.0000

Data Index: 456078 (C4 (Web Text))

Max Activating Token Index: 653