Model: SoLU Model: 12 Layers, 6144 Neurons per Layer

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

Neuron 313 in Layer 8

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-12l')

Text #0

Max Range: 4.2376. Min Range: -4.2376

Max Act: 4.2376. Min Act: -0.0000

Data Index: 1095554 (C4 (Web Text))

Max Activating Token Index: 86

Text #1

Max Range: 4.2376. Min Range: -4.2376

Max Act: 3.8472. Min Act: -0.0000

Data Index: 683539 (C4 (Web Text))

Max Activating Token Index: 359

Text #2

Max Range: 4.2376. Min Range: -4.2376

Max Act: 3.9435. Min Act: -0.0000

Data Index: 444924 (C4 (Web Text))

Max Activating Token Index: 708

Text #3

Max Range: 4.2376. Min Range: -4.2376

Max Act: 3.7083. Min Act: -0.0000

Data Index: 594630 (C4 (Web Text))

Max Activating Token Index: 46

Text #4

Max Range: 4.2376. Min Range: -4.2376

Max Act: 3.5296. Min Act: -0.0000

Data Index: 728095 (C4 (Web Text))

Max Activating Token Index: 406

Text #5

Max Range: 4.2376. Min Range: -4.2376

Max Act: 3.2561. Min Act: -0.0000

Data Index: 1441402 (Python Code)

Max Activating Token Index: 700

Text #6

Max Range: 4.2376. Min Range: -4.2376

Max Act: 3.1911. Min Act: -0.0000

Data Index: 48313 (C4 (Web Text))

Max Activating Token Index: 51

Text #7

Max Range: 4.2376. Min Range: -4.2376

Max Act: 3.1782. Min Act: -0.0000

Data Index: 1341268 (C4 (Web Text))

Max Activating Token Index: 373

Text #8

Max Range: 4.2376. Min Range: -4.2376

Max Act: 3.1211. Min Act: -0.0000

Data Index: 261360 (C4 (Web Text))

Max Activating Token Index: 64

Text #9

Max Range: 4.2376. Min Range: -4.2376

Max Act: 2.8633. Min Act: -0.0000

Data Index: 476468 (C4 (Web Text))

Max Activating Token Index: 599

Text #10

Max Range: 4.2376. Min Range: -4.2376

Max Act: 2.9822. Min Act: -0.0000

Data Index: 349115 (C4 (Web Text))

Max Activating Token Index: 805

Text #11

Max Range: 4.2376. Min Range: -4.2376

Max Act: 2.9538. Min Act: -0.0000

Data Index: 799939 (C4 (Web Text))

Max Activating Token Index: 282

Text #12

Max Range: 4.2376. Min Range: -4.2376

Max Act: 3.0696. Min Act: -0.0000

Data Index: 1110944 (C4 (Web Text))

Max Activating Token Index: 163

Text #13

Max Range: 4.2376. Min Range: -4.2376

Max Act: 2.8962. Min Act: -0.0000

Data Index: 1028519 (C4 (Web Text))

Max Activating Token Index: 624

Text #14

Max Range: 4.2376. Min Range: -4.2376

Max Act: 2.7523. Min Act: -0.0000

Data Index: 247956 (C4 (Web Text))

Max Activating Token Index: 149

Text #15

Max Range: 4.2376. Min Range: -4.2376

Max Act: 2.7761. Min Act: -0.0000

Data Index: 1195368 (C4 (Web Text))

Max Activating Token Index: 81

Text #16

Max Range: 4.2376. Min Range: -4.2376

Max Act: 2.4889. Min Act: -0.0000

Data Index: 428351 (C4 (Web Text))

Max Activating Token Index: 650

Text #17

Max Range: 4.2376. Min Range: -4.2376

Max Act: 2.4453. Min Act: -0.0000

Data Index: 108482 (C4 (Web Text))

Max Activating Token Index: 984

Text #18

Max Range: 4.2376. Min Range: -4.2376

Max Act: 2.4383. Min Act: -0.0000

Data Index: 1523187 (Python Code)

Max Activating Token Index: 447

Text #19

Max Range: 4.2376. Min Range: -4.2376

Max Act: 2.3260. Min Act: -0.0000

Data Index: 1596604 (Python Code)

Max Activating Token Index: 123