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

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

Neuron 3277 in Layer 7

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.9577. Min Range: -0.9577

Max Act: 0.9479. Min Act: -0.0000

Data Index: 104518 (C4 (Web Text))

Max Activating Token Index: 848

Text #1

Max Range: 0.9577. Min Range: -0.9577

Max Act: 0.9577. Min Act: -0.0000

Data Index: 34361 (C4 (Web Text))

Max Activating Token Index: 589

Text #2

Max Range: 0.9577. Min Range: -0.9577

Max Act: 0.8424. Min Act: -0.0001

Data Index: 1145521 (C4 (Web Text))

Max Activating Token Index: 227

Text #3

Max Range: 0.9577. Min Range: -0.9577

Max Act: 0.8010. Min Act: -0.0000

Data Index: 130243 (C4 (Web Text))

Max Activating Token Index: 993

Text #4

Max Range: 0.9577. Min Range: -0.9577

Max Act: 0.8476. Min Act: -0.0000

Data Index: 685529 (C4 (Web Text))

Max Activating Token Index: 164

Text #5

Max Range: 0.9577. Min Range: -0.9577

Max Act: 0.8711. Min Act: -0.0000

Data Index: 1026639 (C4 (Web Text))

Max Activating Token Index: 462

Text #6

Max Range: 0.9577. Min Range: -0.9577

Max Act: 0.8411. Min Act: -0.0000

Data Index: 138185 (C4 (Web Text))

Max Activating Token Index: 570

Text #7

Max Range: 0.9577. Min Range: -0.9577

Max Act: 0.8263. Min Act: -0.0000

Data Index: 1000826 (C4 (Web Text))

Max Activating Token Index: 200

Text #8

Max Range: 0.9577. Min Range: -0.9577

Max Act: 0.7873. Min Act: -0.0001

Data Index: 821337 (C4 (Web Text))

Max Activating Token Index: 827

Text #9

Max Range: 0.9577. Min Range: -0.9577

Max Act: 0.7759. Min Act: -0.0001

Data Index: 1165107 (C4 (Web Text))

Max Activating Token Index: 49

Text #10

Max Range: 0.9577. Min Range: -0.9577

Max Act: 0.7352. Min Act: -0.0000

Data Index: 820689 (C4 (Web Text))

Max Activating Token Index: 329

Text #11

Max Range: 0.9577. Min Range: -0.9577

Max Act: 0.7435. Min Act: -0.0000

Data Index: 1010835 (C4 (Web Text))

Max Activating Token Index: 291

Text #12

Max Range: 0.9577. Min Range: -0.9577

Max Act: 0.7227. Min Act: -0.0000

Data Index: 371266 (C4 (Web Text))

Max Activating Token Index: 141

Text #13

Max Range: 0.9577. Min Range: -0.9577

Max Act: 0.6938. Min Act: -0.0001

Data Index: 418150 (C4 (Web Text))

Max Activating Token Index: 597

Text #14

Max Range: 0.9577. Min Range: -0.9577

Max Act: 0.6938. Min Act: -0.0000

Data Index: 1530882 (Python Code)

Max Activating Token Index: 126

Text #15

Max Range: 0.9577. Min Range: -0.9577

Max Act: 0.7188. Min Act: -0.0001

Data Index: 519508 (C4 (Web Text))

Max Activating Token Index: 889

Text #16

Max Range: 0.9577. Min Range: -0.9577

Max Act: 0.6594. Min Act: -0.0000

Data Index: 447932 (C4 (Web Text))

Max Activating Token Index: 212

Text #17

Max Range: 0.9577. Min Range: -0.9577

Max Act: 0.6509. Min Act: -0.0001

Data Index: 578851 (C4 (Web Text))

Max Activating Token Index: 87

Text #18

Max Range: 0.9577. Min Range: -0.9577

Max Act: 0.6553. Min Act: 0.0001

Data Index: 710086 (C4 (Web Text))

Max Activating Token Index: 118

Text #19

Max Range: 0.9577. Min Range: -0.9577

Max Act: 0.6134. Min Act: -0.0000

Data Index: 1206765 (C4 (Web Text))

Max Activating Token Index: 185