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

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

Neuron 1051 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: 0.8956. Min Range: -0.8956

Max Act: 0.8956. Min Act: -0.0001

Data Index: 971862 (C4 (Web Text))

Max Activating Token Index: 811

Text #1

Max Range: 0.8956. Min Range: -0.8956

Max Act: 0.8918. Min Act: -0.0001

Data Index: 1167931 (C4 (Web Text))

Max Activating Token Index: 129

Text #2

Max Range: 0.8956. Min Range: -0.8956

Max Act: 0.8623. Min Act: -0.0001

Data Index: 1266052 (C4 (Web Text))

Max Activating Token Index: 140

Text #3

Max Range: 0.8956. Min Range: -0.8956

Max Act: 0.8352. Min Act: -0.0001

Data Index: 114685 (C4 (Web Text))

Max Activating Token Index: 382

Text #4

Max Range: 0.8956. Min Range: -0.8956

Max Act: 0.8577. Min Act: -0.0001

Data Index: 835786 (C4 (Web Text))

Max Activating Token Index: 528

Text #5

Max Range: 0.8956. Min Range: -0.8956

Max Act: 0.8862. Min Act: -0.0001

Data Index: 359267 (C4 (Web Text))

Max Activating Token Index: 398

Text #6

Max Range: 0.8956. Min Range: -0.8956

Max Act: 0.8487. Min Act: -0.0001

Data Index: 480339 (C4 (Web Text))

Max Activating Token Index: 368

Text #7

Max Range: 0.8956. Min Range: -0.8956

Max Act: 0.8729. Min Act: -0.0001

Data Index: 224103 (C4 (Web Text))

Max Activating Token Index: 803

Text #8

Max Range: 0.8956. Min Range: -0.8956

Max Act: 0.8277. Min Act: -0.0001

Data Index: 205705 (C4 (Web Text))

Max Activating Token Index: 931

Text #9

Max Range: 0.8956. Min Range: -0.8956

Max Act: 0.8502. Min Act: -0.0001

Data Index: 1123078 (C4 (Web Text))

Max Activating Token Index: 158

Text #10

Max Range: 0.8956. Min Range: -0.8956

Max Act: 0.8198. Min Act: -0.0001

Data Index: 109359 (C4 (Web Text))

Max Activating Token Index: 567

Text #11

Max Range: 0.8956. Min Range: -0.8956

Max Act: 0.8655. Min Act: -0.0001

Data Index: 1301742 (C4 (Web Text))

Max Activating Token Index: 935

Text #12

Max Range: 0.8956. Min Range: -0.8956

Max Act: 0.8408. Min Act: -0.0001

Data Index: 136901 (C4 (Web Text))

Max Activating Token Index: 907

Text #13

Max Range: 0.8956. Min Range: -0.8956

Max Act: 0.8246. Min Act: -0.0001

Data Index: 194422 (C4 (Web Text))

Max Activating Token Index: 724

Text #14

Max Range: 0.8956. Min Range: -0.8956

Max Act: 0.8501. Min Act: -0.0001

Data Index: 1367560 (Python Code)

Max Activating Token Index: 575

Text #15

Max Range: 0.8956. Min Range: -0.8956

Max Act: 0.8213. Min Act: -0.0001

Data Index: 550131 (C4 (Web Text))

Max Activating Token Index: 260

Text #16

Max Range: 0.8956. Min Range: -0.8956

Max Act: 0.8128. Min Act: -0.0001

Data Index: 440250 (C4 (Web Text))

Max Activating Token Index: 82

Text #17

Max Range: 0.8956. Min Range: -0.8956

Max Act: 0.7943. Min Act: -0.0001

Data Index: 1266740 (C4 (Web Text))

Max Activating Token Index: 488

Text #18

Max Range: 0.8956. Min Range: -0.8956

Max Act: 0.8364. Min Act: -0.0001

Data Index: 186096 (C4 (Web Text))

Max Activating Token Index: 183

Text #19

Max Range: 0.8956. Min Range: -0.8956

Max Act: 0.7846. Min Act: -0.0001

Data Index: 403843 (C4 (Web Text))

Max Activating Token Index: 394