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

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

Neuron 4031 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: 0.0467. Min Range: -0.0467

Max Act: 0.0467. Min Act: -0.0001

Data Index: 409271 (C4 (Web Text))

Max Activating Token Index: 141

Text #1

Max Range: 0.0467. Min Range: -0.0467

Max Act: 0.0443. Min Act: -0.0001

Data Index: 940577 (C4 (Web Text))

Max Activating Token Index: 113

Text #2

Max Range: 0.0467. Min Range: -0.0467

Max Act: 0.0445. Min Act: -0.0001

Data Index: 36644 (C4 (Web Text))

Max Activating Token Index: 740

Text #3

Max Range: 0.0467. Min Range: -0.0467

Max Act: 0.0442. Min Act: -0.0001

Data Index: 230506 (C4 (Web Text))

Max Activating Token Index: 260

Text #4

Max Range: 0.0467. Min Range: -0.0467

Max Act: 0.0405. Min Act: -0.0001

Data Index: 731957 (C4 (Web Text))

Max Activating Token Index: 857

Text #5

Max Range: 0.0467. Min Range: -0.0467

Max Act: 0.0409. Min Act: -0.0001

Data Index: 1449126 (Python Code)

Max Activating Token Index: 256

Text #6

Max Range: 0.0467. Min Range: -0.0467

Max Act: 0.0406. Min Act: -0.0001

Data Index: 533431 (C4 (Web Text))

Max Activating Token Index: 236

Text #7

Max Range: 0.0467. Min Range: -0.0467

Max Act: 0.0404. Min Act: -0.0001

Data Index: 1155182 (C4 (Web Text))

Max Activating Token Index: 32

Text #8

Max Range: 0.0467. Min Range: -0.0467

Max Act: 0.0378. Min Act: -0.0001

Data Index: 352007 (C4 (Web Text))

Max Activating Token Index: 627

Text #9

Max Range: 0.0467. Min Range: -0.0467

Max Act: 0.0404. Min Act: -0.0001

Data Index: 538116 (C4 (Web Text))

Max Activating Token Index: 54

Text #10

Max Range: 0.0467. Min Range: -0.0467

Max Act: 0.0382. Min Act: -0.0001

Data Index: 115732 (C4 (Web Text))

Max Activating Token Index: 544

Text #11

Max Range: 0.0467. Min Range: -0.0467

Max Act: 0.0381. Min Act: -0.0001

Data Index: 1580417 (Python Code)

Max Activating Token Index: 345

Text #12

Max Range: 0.0467. Min Range: -0.0467

Max Act: 0.0377. Min Act: -0.0001

Data Index: 330788 (C4 (Web Text))

Max Activating Token Index: 429

Text #13

Max Range: 0.0467. Min Range: -0.0467

Max Act: 0.0374. Min Act: -0.0001

Data Index: 1561924 (Python Code)

Max Activating Token Index: 6

Text #14

Max Range: 0.0467. Min Range: -0.0467

Max Act: 0.0360. Min Act: -0.0001

Data Index: 1399031 (Python Code)

Max Activating Token Index: 994

Text #15

Max Range: 0.0467. Min Range: -0.0467

Max Act: 0.0369. Min Act: -0.0001

Data Index: 1611053 (Python Code)

Max Activating Token Index: 32

Text #16

Max Range: 0.0467. Min Range: -0.0467

Max Act: 0.0353. Min Act: -0.0001

Data Index: 1515428 (Python Code)

Max Activating Token Index: 833

Text #17

Max Range: 0.0467. Min Range: -0.0467

Max Act: 0.0360. Min Act: -0.0001

Data Index: 5224 (C4 (Web Text))

Max Activating Token Index: 754

Text #18

Max Range: 0.0467. Min Range: -0.0467

Max Act: 0.0371. Min Act: -0.0001

Data Index: 882485 (C4 (Web Text))

Max Activating Token Index: 684

Text #19

Max Range: 0.0467. Min Range: -0.0467

Max Act: 0.0381. Min Act: -0.0001

Data Index: 718131 (C4 (Web Text))

Max Activating Token Index: 665