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

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

Neuron 1712 in Layer 4

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-8l')

Text #0

Max Range: 0.5467. Min Range: -0.5467

Max Act: 0.5467. Min Act: -0.0001

Data Index: 844770 (C4 (Web Text))

Max Activating Token Index: 951

Text #1

Max Range: 0.5467. Min Range: -0.5467

Max Act: 0.5274. Min Act: -0.0001

Data Index: 224525 (C4 (Web Text))

Max Activating Token Index: 323

Text #2

Max Range: 0.5467. Min Range: -0.5467

Max Act: 0.4889. Min Act: -0.0001

Data Index: 787919 (C4 (Web Text))

Max Activating Token Index: 860

Text #3

Max Range: 0.5467. Min Range: -0.5467

Max Act: 0.5038. Min Act: -0.0001

Data Index: 791114 (C4 (Web Text))

Max Activating Token Index: 389

Text #4

Max Range: 0.5467. Min Range: -0.5467

Max Act: 0.5015. Min Act: -0.0001

Data Index: 528377 (C4 (Web Text))

Max Activating Token Index: 22

Text #5

Max Range: 0.5467. Min Range: -0.5467

Max Act: 0.4995. Min Act: -0.0001

Data Index: 1163515 (C4 (Web Text))

Max Activating Token Index: 726

Text #6

Max Range: 0.5467. Min Range: -0.5467

Max Act: 0.4597. Min Act: -0.0001

Data Index: 13414 (C4 (Web Text))

Max Activating Token Index: 86

Text #7

Max Range: 0.5467. Min Range: -0.5467

Max Act: 0.4810. Min Act: -0.0001

Data Index: 1038013 (C4 (Web Text))

Max Activating Token Index: 190

Text #8

Max Range: 0.5467. Min Range: -0.5467

Max Act: 0.4640. Min Act: -0.0001

Data Index: 924240 (C4 (Web Text))

Max Activating Token Index: 531

Text #9

Max Range: 0.5467. Min Range: -0.5467

Max Act: 0.4733. Min Act: -0.0001

Data Index: 341613 (C4 (Web Text))

Max Activating Token Index: 15

Text #10

Max Range: 0.5467. Min Range: -0.5467

Max Act: 0.4255. Min Act: -0.0001

Data Index: 984306 (C4 (Web Text))

Max Activating Token Index: 315

Text #11

Max Range: 0.5467. Min Range: -0.5467

Max Act: 0.4551. Min Act: -0.0001

Data Index: 324303 (C4 (Web Text))

Max Activating Token Index: 24

Text #12

Max Range: 0.5467. Min Range: -0.5467

Max Act: 0.4832. Min Act: -0.0001

Data Index: 575487 (C4 (Web Text))

Max Activating Token Index: 796

Text #13

Max Range: 0.5467. Min Range: -0.5467

Max Act: 0.4520. Min Act: -0.0001

Data Index: 916938 (C4 (Web Text))

Max Activating Token Index: 20

Text #14

Max Range: 0.5467. Min Range: -0.5467

Max Act: 0.4325. Min Act: -0.0001

Data Index: 35163 (C4 (Web Text))

Max Activating Token Index: 172

Text #15

Max Range: 0.5467. Min Range: -0.5467

Max Act: 0.4276. Min Act: -0.0001

Data Index: 1331439 (C4 (Web Text))

Max Activating Token Index: 405

Text #16

Max Range: 0.5467. Min Range: -0.5467

Max Act: 0.4182. Min Act: -0.0001

Data Index: 407939 (C4 (Web Text))

Max Activating Token Index: 909

Text #17

Max Range: 0.5467. Min Range: -0.5467

Max Act: 0.4230. Min Act: -0.0001

Data Index: 283728 (C4 (Web Text))

Max Activating Token Index: 463

Text #18

Max Range: 0.5467. Min Range: -0.5467

Max Act: 0.4090. Min Act: -0.0001

Data Index: 915614 (C4 (Web Text))

Max Activating Token Index: 708

Text #19

Max Range: 0.5467. Min Range: -0.5467

Max Act: 0.4227. Min Act: -0.0001

Data Index: 1358291 (C4 (Web Text))

Max Activating Token Index: 472