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

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

Neuron 3227 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: 0.1847. Min Range: -0.1847

Max Act: 0.1834. Min Act: -0.0000

Data Index: 928537 (C4 (Web Text))

Max Activating Token Index: 180

Text #1

Max Range: 0.1847. Min Range: -0.1847

Max Act: 0.1774. Min Act: -0.0000

Data Index: 1641696 (Python Code)

Max Activating Token Index: 617

Text #2

Max Range: 0.1847. Min Range: -0.1847

Max Act: 0.1804. Min Act: -0.0000

Data Index: 735056 (C4 (Web Text))

Max Activating Token Index: 204

Text #3

Max Range: 0.1847. Min Range: -0.1847

Max Act: 0.1847. Min Act: -0.0001

Data Index: 1643752 (Python Code)

Max Activating Token Index: 329

Text #4

Max Range: 0.1847. Min Range: -0.1847

Max Act: 0.1662. Min Act: -0.0000

Data Index: 1540638 (Python Code)

Max Activating Token Index: 200

Text #5

Max Range: 0.1847. Min Range: -0.1847

Max Act: 0.1692. Min Act: -0.0000

Data Index: 681709 (C4 (Web Text))

Max Activating Token Index: 453

Text #6

Max Range: 0.1847. Min Range: -0.1847

Max Act: 0.1591. Min Act: -0.0000

Data Index: 538541 (C4 (Web Text))

Max Activating Token Index: 776

Text #7

Max Range: 0.1847. Min Range: -0.1847

Max Act: 0.1480. Min Act: -0.0000

Data Index: 1432824 (Python Code)

Max Activating Token Index: 769

Text #8

Max Range: 0.1847. Min Range: -0.1847

Max Act: 0.1504. Min Act: -0.0000

Data Index: 35943 (C4 (Web Text))

Max Activating Token Index: 958

Text #9

Max Range: 0.1847. Min Range: -0.1847

Max Act: 0.1490. Min Act: -0.0000

Data Index: 293284 (C4 (Web Text))

Max Activating Token Index: 483

Text #10

Max Range: 0.1847. Min Range: -0.1847

Max Act: 0.1441. Min Act: -0.0000

Data Index: 959883 (C4 (Web Text))

Max Activating Token Index: 438

Text #11

Max Range: 0.1847. Min Range: -0.1847

Max Act: 0.1491. Min Act: 0.0000

Data Index: 1521408 (Python Code)

Max Activating Token Index: 789

Text #12

Max Range: 0.1847. Min Range: -0.1847

Max Act: 0.1430. Min Act: -0.0000

Data Index: 857222 (C4 (Web Text))

Max Activating Token Index: 355

Text #13

Max Range: 0.1847. Min Range: -0.1847

Max Act: 0.1526. Min Act: -0.0000

Data Index: 1011038 (C4 (Web Text))

Max Activating Token Index: 839

Text #14

Max Range: 0.1847. Min Range: -0.1847

Max Act: 0.1411. Min Act: -0.0000

Data Index: 1390554 (Python Code)

Max Activating Token Index: 689

Text #15

Max Range: 0.1847. Min Range: -0.1847

Max Act: 0.1477. Min Act: -0.0000

Data Index: 1123633 (C4 (Web Text))

Max Activating Token Index: 675

Text #16

Max Range: 0.1847. Min Range: -0.1847

Max Act: 0.1456. Min Act: -0.0000

Data Index: 1444956 (Python Code)

Max Activating Token Index: 989

Text #17

Max Range: 0.1847. Min Range: -0.1847

Max Act: 0.1465. Min Act: -0.0000

Data Index: 923285 (C4 (Web Text))

Max Activating Token Index: 674

Text #18

Max Range: 0.1847. Min Range: -0.1847

Max Act: 0.1318. Min Act: -0.0000

Data Index: 1070846 (C4 (Web Text))

Max Activating Token Index: 783

Text #19

Max Range: 0.1847. Min Range: -0.1847

Max Act: 0.1343. Min Act: -0.0000

Data Index: 337292 (C4 (Web Text))

Max Activating Token Index: 263