Model: SoLU Model: 8 Layers, 4096 Neurons per Layer

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

Neuron 2556 in Layer 7

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-8l')

Text #0

Max Range: 0.5783. Min Range: -0.5783

Max Act: 0.5783. Min Act: -0.0001

Data Index: 1248595 (C4 (Web Text))

Max Activating Token Index: 970

Text #1

Max Range: 0.5783. Min Range: -0.5783

Max Act: 0.5325. Min Act: -0.0000

Data Index: 141955 (C4 (Web Text))

Max Activating Token Index: 584

Text #2

Max Range: 0.5783. Min Range: -0.5783

Max Act: 0.5371. Min Act: -0.0000

Data Index: 1598940 (Python Code)

Max Activating Token Index: 68

Text #3

Max Range: 0.5783. Min Range: -0.5783

Max Act: 0.5162. Min Act: -0.0000

Data Index: 1161959 (C4 (Web Text))

Max Activating Token Index: 858

Text #4

Max Range: 0.5783. Min Range: -0.5783

Max Act: 0.4794. Min Act: -0.0000

Data Index: 1038420 (C4 (Web Text))

Max Activating Token Index: 480

Text #5

Max Range: 0.5783. Min Range: -0.5783

Max Act: 0.4994. Min Act: -0.0001

Data Index: 80864 (C4 (Web Text))

Max Activating Token Index: 1003

Text #6

Max Range: 0.5783. Min Range: -0.5783

Max Act: 0.4816. Min Act: -0.0000

Data Index: 420668 (C4 (Web Text))

Max Activating Token Index: 664

Text #7

Max Range: 0.5783. Min Range: -0.5783

Max Act: 0.5024. Min Act: -0.0001

Data Index: 1143961 (C4 (Web Text))

Max Activating Token Index: 831

Text #8

Max Range: 0.5783. Min Range: -0.5783

Max Act: 0.4807. Min Act: -0.0000

Data Index: 804271 (C4 (Web Text))

Max Activating Token Index: 270

Text #9

Max Range: 0.5783. Min Range: -0.5783

Max Act: 0.4437. Min Act: -0.0000

Data Index: 1418297 (Python Code)

Max Activating Token Index: 827

Text #10

Max Range: 0.5783. Min Range: -0.5783

Max Act: 0.4547. Min Act: -0.0000

Data Index: 621309 (C4 (Web Text))

Max Activating Token Index: 360

Text #11

Max Range: 0.5783. Min Range: -0.5783

Max Act: 0.4391. Min Act: -0.0000

Data Index: 1458396 (Python Code)

Max Activating Token Index: 84

Text #12

Max Range: 0.5783. Min Range: -0.5783

Max Act: 0.4386. Min Act: -0.0001

Data Index: 350714 (C4 (Web Text))

Max Activating Token Index: 473

Text #13

Max Range: 0.5783. Min Range: -0.5783

Max Act: 0.4229. Min Act: -0.0000

Data Index: 1612126 (Python Code)

Max Activating Token Index: 298

Text #14

Max Range: 0.5783. Min Range: -0.5783

Max Act: 0.4400. Min Act: -0.0000

Data Index: 591310 (C4 (Web Text))

Max Activating Token Index: 971

Text #15

Max Range: 0.5783. Min Range: -0.5783

Max Act: 0.4131. Min Act: -0.0001

Data Index: 1589957 (Python Code)

Max Activating Token Index: 645

Text #16

Max Range: 0.5783. Min Range: -0.5783

Max Act: 0.4164. Min Act: -0.0001

Data Index: 1433217 (Python Code)

Max Activating Token Index: 476

Text #17

Max Range: 0.5783. Min Range: -0.5783

Max Act: 0.4039. Min Act: -0.0000

Data Index: 1242603 (C4 (Web Text))

Max Activating Token Index: 318

Text #18

Max Range: 0.5783. Min Range: -0.5783

Max Act: 0.4158. Min Act: -0.0000

Data Index: 766508 (C4 (Web Text))

Max Activating Token Index: 1009

Text #19

Max Range: 0.5783. Min Range: -0.5783

Max Act: 0.4124. Min Act: 0.0000

Data Index: 694667 (C4 (Web Text))

Max Activating Token Index: 495