Model: SoLU Model: 10 Layers, 5120 Neurons per Layer

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

Neuron 442 in Layer 0

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-10l')

Text #0

Max Range: 0.3794. Min Range: -0.3794

Max Act: 0.3794. Min Act: -0.0002

Data Index: 259980 (C4 (Web Text))

Max Activating Token Index: 746

Text #1

Max Range: 0.3794. Min Range: -0.3794

Max Act: 0.3658. Min Act: -0.0002

Data Index: 177110 (C4 (Web Text))

Max Activating Token Index: 361

Text #2

Max Range: 0.3794. Min Range: -0.3794

Max Act: 0.3422. Min Act: -0.0002

Data Index: 676284 (C4 (Web Text))

Max Activating Token Index: 418

Text #3

Max Range: 0.3794. Min Range: -0.3794

Max Act: 0.3373. Min Act: -0.0002

Data Index: 1542154 (Python Code)

Max Activating Token Index: 626

Text #4

Max Range: 0.3794. Min Range: -0.3794

Max Act: 0.3329. Min Act: -0.0002

Data Index: 365329 (C4 (Web Text))

Max Activating Token Index: 884

Text #5

Max Range: 0.3794. Min Range: -0.3794

Max Act: 0.3258. Min Act: -0.0002

Data Index: 930116 (C4 (Web Text))

Max Activating Token Index: 503

Text #6

Max Range: 0.3794. Min Range: -0.3794

Max Act: 0.3140. Min Act: -0.0002

Data Index: 832815 (C4 (Web Text))

Max Activating Token Index: 975

Text #7

Max Range: 0.3794. Min Range: -0.3794

Max Act: 0.3074. Min Act: -0.0002

Data Index: 10699 (C4 (Web Text))

Max Activating Token Index: 184

Text #8

Max Range: 0.3794. Min Range: -0.3794

Max Act: 0.3031. Min Act: -0.0002

Data Index: 395641 (C4 (Web Text))

Max Activating Token Index: 851

Text #9

Max Range: 0.3794. Min Range: -0.3794

Max Act: 0.3002. Min Act: -0.0002

Data Index: 502056 (C4 (Web Text))

Max Activating Token Index: 85

Text #10

Max Range: 0.3794. Min Range: -0.3794

Max Act: 0.2851. Min Act: -0.0002

Data Index: 1400584 (Python Code)

Max Activating Token Index: 146

Text #11

Max Range: 0.3794. Min Range: -0.3794

Max Act: 0.2689. Min Act: -0.0002

Data Index: 881568 (C4 (Web Text))

Max Activating Token Index: 385

Text #12

Max Range: 0.3794. Min Range: -0.3794

Max Act: 0.2606. Min Act: -0.0002

Data Index: 157271 (C4 (Web Text))

Max Activating Token Index: 771

Text #13

Max Range: 0.3794. Min Range: -0.3794

Max Act: 0.2696. Min Act: -0.0002

Data Index: 1483444 (Python Code)

Max Activating Token Index: 946

Text #14

Max Range: 0.3794. Min Range: -0.3794

Max Act: 0.2570. Min Act: -0.0002

Data Index: 331096 (C4 (Web Text))

Max Activating Token Index: 1002

Text #15

Max Range: 0.3794. Min Range: -0.3794

Max Act: 0.2559. Min Act: -0.0002

Data Index: 1231420 (C4 (Web Text))

Max Activating Token Index: 966

Text #16

Max Range: 0.3794. Min Range: -0.3794

Max Act: 0.2564. Min Act: -0.0002

Data Index: 618679 (C4 (Web Text))

Max Activating Token Index: 630

Text #17

Max Range: 0.3794. Min Range: -0.3794

Max Act: 0.2500. Min Act: -0.0002

Data Index: 558253 (C4 (Web Text))

Max Activating Token Index: 606

Text #18

Max Range: 0.3794. Min Range: -0.3794

Max Act: 0.2497. Min Act: -0.0002

Data Index: 946112 (C4 (Web Text))

Max Activating Token Index: 443

Text #19

Max Range: 0.3794. Min Range: -0.3794

Max Act: 0.2497. Min Act: -0.0002

Data Index: 544257 (C4 (Web Text))

Max Activating Token Index: 561