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

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

Neuron 2867 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: 5.6335. Min Range: -5.6335

Max Act: 5.6335. Min Act: -0.0000

Data Index: 66955 (C4 (Web Text))

Max Activating Token Index: 520

Text #1

Max Range: 5.6335. Min Range: -5.6335

Max Act: 5.0046. Min Act: -0.0000

Data Index: 1444023 (Python Code)

Max Activating Token Index: 670

Text #2

Max Range: 5.6335. Min Range: -5.6335

Max Act: 4.9493. Min Act: -0.0000

Data Index: 1283937 (C4 (Web Text))

Max Activating Token Index: 36

Text #3

Max Range: 5.6335. Min Range: -5.6335

Max Act: 4.7225. Min Act: -0.0000

Data Index: 1524766 (Python Code)

Max Activating Token Index: 578

Text #4

Max Range: 5.6335. Min Range: -5.6335

Max Act: 4.5390. Min Act: -0.0000

Data Index: 109165 (C4 (Web Text))

Max Activating Token Index: 167

Text #5

Max Range: 5.6335. Min Range: -5.6335

Max Act: 5.0726. Min Act: -0.0000

Data Index: 1603956 (Python Code)

Max Activating Token Index: 804

Text #6

Max Range: 5.6335. Min Range: -5.6335

Max Act: 4.2724. Min Act: -0.0000

Data Index: 1332255 (C4 (Web Text))

Max Activating Token Index: 566

Text #7

Max Range: 5.6335. Min Range: -5.6335

Max Act: 4.8239. Min Act: -0.0000

Data Index: 515 (C4 (Web Text))

Max Activating Token Index: 592

Text #8

Max Range: 5.6335. Min Range: -5.6335

Max Act: 3.9633. Min Act: -0.0000

Data Index: 1482209 (Python Code)

Max Activating Token Index: 655

Text #9

Max Range: 5.6335. Min Range: -5.6335

Max Act: 3.8950. Min Act: -0.0000

Data Index: 1004174 (C4 (Web Text))

Max Activating Token Index: 344

Text #10

Max Range: 5.6335. Min Range: -5.6335

Max Act: 4.2615. Min Act: -0.0000

Data Index: 776144 (C4 (Web Text))

Max Activating Token Index: 242

Text #11

Max Range: 5.6335. Min Range: -5.6335

Max Act: 3.9165. Min Act: -0.0000

Data Index: 848095 (C4 (Web Text))

Max Activating Token Index: 220

Text #12

Max Range: 5.6335. Min Range: -5.6335

Max Act: 4.2394. Min Act: -0.0000

Data Index: 211949 (C4 (Web Text))

Max Activating Token Index: 373

Text #13

Max Range: 5.6335. Min Range: -5.6335

Max Act: 4.0529. Min Act: -0.0000

Data Index: 154101 (C4 (Web Text))

Max Activating Token Index: 131

Text #14

Max Range: 5.6335. Min Range: -5.6335

Max Act: 3.9976. Min Act: -0.0000

Data Index: 1469581 (Python Code)

Max Activating Token Index: 772

Text #15

Max Range: 5.6335. Min Range: -5.6335

Max Act: 4.0658. Min Act: -0.0000

Data Index: 136006 (C4 (Web Text))

Max Activating Token Index: 506

Text #16

Max Range: 5.6335. Min Range: -5.6335

Max Act: 4.2343. Min Act: -0.0000

Data Index: 1292345 (C4 (Web Text))

Max Activating Token Index: 548

Text #17

Max Range: 5.6335. Min Range: -5.6335

Max Act: 3.8892. Min Act: -0.0000

Data Index: 1452016 (Python Code)

Max Activating Token Index: 236

Text #18

Max Range: 5.6335. Min Range: -5.6335

Max Act: 3.2011. Min Act: -0.0000

Data Index: 1629482 (Python Code)

Max Activating Token Index: 287

Text #19

Max Range: 5.6335. Min Range: -5.6335

Max Act: 3.9830. Min Act: -0.0000

Data Index: 127749 (C4 (Web Text))

Max Activating Token Index: 27