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

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

Neuron 608 in Layer 3

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.1567. Min Range: -0.1567

Max Act: 0.1567. Min Act: -0.0001

Data Index: 1389172 (Python Code)

Max Activating Token Index: 395

Text #1

Max Range: 0.1567. Min Range: -0.1567

Max Act: 0.1273. Min Act: -0.0001

Data Index: 827591 (C4 (Web Text))

Max Activating Token Index: 557

Text #2

Max Range: 0.1567. Min Range: -0.1567

Max Act: 0.1538. Min Act: -0.0001

Data Index: 1060168 (C4 (Web Text))

Max Activating Token Index: 16

Text #3

Max Range: 0.1567. Min Range: -0.1567

Max Act: 0.1500. Min Act: -0.0001

Data Index: 936395 (C4 (Web Text))

Max Activating Token Index: 642

Text #4

Max Range: 0.1567. Min Range: -0.1567

Max Act: 0.1461. Min Act: -0.0001

Data Index: 144722 (C4 (Web Text))

Max Activating Token Index: 115

Text #5

Max Range: 0.1567. Min Range: -0.1567

Max Act: 0.1300. Min Act: -0.0001

Data Index: 590112 (C4 (Web Text))

Max Activating Token Index: 194

Text #6

Max Range: 0.1567. Min Range: -0.1567

Max Act: 0.1429. Min Act: -0.0001

Data Index: 875425 (C4 (Web Text))

Max Activating Token Index: 130

Text #7

Max Range: 0.1567. Min Range: -0.1567

Max Act: 0.1301. Min Act: -0.0001

Data Index: 304874 (C4 (Web Text))

Max Activating Token Index: 892

Text #8

Max Range: 0.1567. Min Range: -0.1567

Max Act: 0.1411. Min Act: -0.0001

Data Index: 646169 (C4 (Web Text))

Max Activating Token Index: 71

Text #9

Max Range: 0.1567. Min Range: -0.1567

Max Act: 0.1245. Min Act: -0.0001

Data Index: 1255849 (C4 (Web Text))

Max Activating Token Index: 714

Text #10

Max Range: 0.1567. Min Range: -0.1567

Max Act: 0.1323. Min Act: -0.0001

Data Index: 741604 (C4 (Web Text))

Max Activating Token Index: 411

Text #11

Max Range: 0.1567. Min Range: -0.1567

Max Act: 0.1333. Min Act: -0.0001

Data Index: 1062485 (C4 (Web Text))

Max Activating Token Index: 89

Text #12

Max Range: 0.1567. Min Range: -0.1567

Max Act: 0.1358. Min Act: -0.0001

Data Index: 764988 (C4 (Web Text))

Max Activating Token Index: 29

Text #13

Max Range: 0.1567. Min Range: -0.1567

Max Act: 0.1293. Min Act: -0.0001

Data Index: 264538 (C4 (Web Text))

Max Activating Token Index: 347

Text #14

Max Range: 0.1567. Min Range: -0.1567

Max Act: 0.1254. Min Act: -0.0001

Data Index: 278358 (C4 (Web Text))

Max Activating Token Index: 536

Text #15

Max Range: 0.1567. Min Range: -0.1567

Max Act: 0.1284. Min Act: -0.0001

Data Index: 550045 (C4 (Web Text))

Max Activating Token Index: 736

Text #16

Max Range: 0.1567. Min Range: -0.1567

Max Act: 0.1185. Min Act: -0.0001

Data Index: 68459 (C4 (Web Text))

Max Activating Token Index: 218

Text #17

Max Range: 0.1567. Min Range: -0.1567

Max Act: 0.1146. Min Act: -0.0001

Data Index: 749461 (C4 (Web Text))

Max Activating Token Index: 814

Text #18

Max Range: 0.1567. Min Range: -0.1567

Max Act: 0.1270. Min Act: -0.0001

Data Index: 1548178 (Python Code)

Max Activating Token Index: 664

Text #19

Max Range: 0.1567. Min Range: -0.1567

Max Act: 0.1204. Min Act: -0.0001

Data Index: 333785 (C4 (Web Text))

Max Activating Token Index: 934