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

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

Neuron 1694 in Layer 7

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 1.6893. Min Range: -1.6893

Max Act: 1.6893. Min Act: -0.0000

Data Index: 772367 (C4 (Web Text))

Max Activating Token Index: 219

Text #1

Max Range: 1.6893. Min Range: -1.6893

Max Act: 1.5466. Min Act: -0.0000

Data Index: 859541 (C4 (Web Text))

Max Activating Token Index: 107

Text #2

Max Range: 1.6893. Min Range: -1.6893

Max Act: 1.4797. Min Act: -0.0000

Data Index: 1558056 (Python Code)

Max Activating Token Index: 479

Text #3

Max Range: 1.6893. Min Range: -1.6893

Max Act: 1.3031. Min Act: -0.0000

Data Index: 454303 (C4 (Web Text))

Max Activating Token Index: 290

Text #4

Max Range: 1.6893. Min Range: -1.6893

Max Act: 1.3598. Min Act: -0.0000

Data Index: 1373117 (Python Code)

Max Activating Token Index: 664

Text #5

Max Range: 1.6893. Min Range: -1.6893

Max Act: 1.2930. Min Act: -0.0000

Data Index: 1203171 (C4 (Web Text))

Max Activating Token Index: 818

Text #6

Max Range: 1.6893. Min Range: -1.6893

Max Act: 1.2719. Min Act: -0.0000

Data Index: 408560 (C4 (Web Text))

Max Activating Token Index: 715

Text #7

Max Range: 1.6893. Min Range: -1.6893

Max Act: 1.2699. Min Act: -0.0000

Data Index: 1347535 (C4 (Web Text))

Max Activating Token Index: 249

Text #8

Max Range: 1.6893. Min Range: -1.6893

Max Act: 1.2866. Min Act: -0.0000

Data Index: 831746 (C4 (Web Text))

Max Activating Token Index: 731

Text #9

Max Range: 1.6893. Min Range: -1.6893

Max Act: 1.2453. Min Act: -0.0000

Data Index: 338674 (C4 (Web Text))

Max Activating Token Index: 66

Text #10

Max Range: 1.6893. Min Range: -1.6893

Max Act: 1.2130. Min Act: -0.0000

Data Index: 360752 (C4 (Web Text))

Max Activating Token Index: 374

Text #11

Max Range: 1.6893. Min Range: -1.6893

Max Act: 1.1933. Min Act: -0.0000

Data Index: 1259812 (C4 (Web Text))

Max Activating Token Index: 144

Text #12

Max Range: 1.6893. Min Range: -1.6893

Max Act: 1.1881. Min Act: -0.0000

Data Index: 1059080 (C4 (Web Text))

Max Activating Token Index: 447

Text #13

Max Range: 1.6893. Min Range: -1.6893

Max Act: 1.1799. Min Act: -0.0000

Data Index: 939883 (C4 (Web Text))

Max Activating Token Index: 16

Text #14

Max Range: 1.6893. Min Range: -1.6893

Max Act: 1.1296. Min Act: -0.0000

Data Index: 77469 (C4 (Web Text))

Max Activating Token Index: 136

Text #15

Max Range: 1.6893. Min Range: -1.6893

Max Act: 1.1079. Min Act: -0.0000

Data Index: 1465711 (Python Code)

Max Activating Token Index: 802

Text #16

Max Range: 1.6893. Min Range: -1.6893

Max Act: 1.1341. Min Act: -0.0000

Data Index: 250647 (C4 (Web Text))

Max Activating Token Index: 975

Text #17

Max Range: 1.6893. Min Range: -1.6893

Max Act: 1.1261. Min Act: -0.0000

Data Index: 410856 (C4 (Web Text))

Max Activating Token Index: 566

Text #18

Max Range: 1.6893. Min Range: -1.6893

Max Act: 1.1723. Min Act: -0.0000

Data Index: 123320 (C4 (Web Text))

Max Activating Token Index: 295

Text #19

Max Range: 1.6893. Min Range: -1.6893

Max Act: 1.0787. Min Act: -0.0000

Data Index: 82474 (C4 (Web Text))

Max Activating Token Index: 38