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

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

Neuron 145 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: 2.5885. Min Range: -2.5885

Max Act: 2.5885. Min Act: -0.0000

Data Index: 94885 (C4 (Web Text))

Max Activating Token Index: 963

Text #1

Max Range: 2.5885. Min Range: -2.5885

Max Act: 2.4452. Min Act: -0.0000

Data Index: 1410145 (Python Code)

Max Activating Token Index: 235

Text #2

Max Range: 2.5885. Min Range: -2.5885

Max Act: 2.4692. Min Act: -0.0000

Data Index: 1379953 (Python Code)

Max Activating Token Index: 344

Text #3

Max Range: 2.5885. Min Range: -2.5885

Max Act: 2.2127. Min Act: -0.0000

Data Index: 941371 (C4 (Web Text))

Max Activating Token Index: 18

Text #4

Max Range: 2.5885. Min Range: -2.5885

Max Act: 2.2514. Min Act: -0.0000

Data Index: 178720 (C4 (Web Text))

Max Activating Token Index: 453

Text #5

Max Range: 2.5885. Min Range: -2.5885

Max Act: 2.1263. Min Act: -0.0000

Data Index: 1450019 (Python Code)

Max Activating Token Index: 264

Text #6

Max Range: 2.5885. Min Range: -2.5885

Max Act: 2.2171. Min Act: -0.0000

Data Index: 1473230 (Python Code)

Max Activating Token Index: 591

Text #7

Max Range: 2.5885. Min Range: -2.5885

Max Act: 2.2029. Min Act: -0.0000

Data Index: 1595376 (Python Code)

Max Activating Token Index: 397

Text #8

Max Range: 2.5885. Min Range: -2.5885

Max Act: 2.0761. Min Act: -0.0000

Data Index: 1362994 (Python Code)

Max Activating Token Index: 145

Text #9

Max Range: 2.5885. Min Range: -2.5885

Max Act: 2.0330. Min Act: -0.0000

Data Index: 240400 (C4 (Web Text))

Max Activating Token Index: 436

Text #10

Max Range: 2.5885. Min Range: -2.5885

Max Act: 1.8872. Min Act: -0.0000

Data Index: 1582390 (Python Code)

Max Activating Token Index: 258

Text #11

Max Range: 2.5885. Min Range: -2.5885

Max Act: 1.9817. Min Act: -0.0000

Data Index: 1400843 (Python Code)

Max Activating Token Index: 246

Text #12

Max Range: 2.5885. Min Range: -2.5885

Max Act: 1.9737. Min Act: -0.0000

Data Index: 1604683 (Python Code)

Max Activating Token Index: 885

Text #13

Max Range: 2.5885. Min Range: -2.5885

Max Act: 1.9613. Min Act: -0.0000

Data Index: 301729 (C4 (Web Text))

Max Activating Token Index: 94

Text #14

Max Range: 2.5885. Min Range: -2.5885

Max Act: 1.9518. Min Act: -0.0000

Data Index: 1478558 (Python Code)

Max Activating Token Index: 1002

Text #15

Max Range: 2.5885. Min Range: -2.5885

Max Act: 1.9548. Min Act: -0.0000

Data Index: 1464500 (Python Code)

Max Activating Token Index: 911

Text #16

Max Range: 2.5885. Min Range: -2.5885

Max Act: 1.8315. Min Act: -0.0000

Data Index: 1432128 (Python Code)

Max Activating Token Index: 483

Text #17

Max Range: 2.5885. Min Range: -2.5885

Max Act: 1.9383. Min Act: -0.0000

Data Index: 864143 (C4 (Web Text))

Max Activating Token Index: 924

Text #18

Max Range: 2.5885. Min Range: -2.5885

Max Act: 1.9431. Min Act: -0.0000

Data Index: 1326087 (C4 (Web Text))

Max Activating Token Index: 441

Text #19

Max Range: 2.5885. Min Range: -2.5885

Max Act: 2.0407. Min Act: -0.0000

Data Index: 1479108 (Python Code)

Max Activating Token Index: 1006