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

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

Neuron 3966 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: 0.5483. Min Range: -0.5483

Max Act: 0.5337. Min Act: -0.0001

Data Index: 335664 (C4 (Web Text))

Max Activating Token Index: 820

Text #1

Max Range: 0.5483. Min Range: -0.5483

Max Act: 0.5483. Min Act: -0.0000

Data Index: 486300 (C4 (Web Text))

Max Activating Token Index: 935

Text #2

Max Range: 0.5483. Min Range: -0.5483

Max Act: 0.5139. Min Act: -0.0001

Data Index: 710919 (C4 (Web Text))

Max Activating Token Index: 368

Text #3

Max Range: 0.5483. Min Range: -0.5483

Max Act: 0.5109. Min Act: -0.0001

Data Index: 928599 (C4 (Web Text))

Max Activating Token Index: 163

Text #4

Max Range: 0.5483. Min Range: -0.5483

Max Act: 0.4955. Min Act: -0.0001

Data Index: 79053 (C4 (Web Text))

Max Activating Token Index: 104

Text #5

Max Range: 0.5483. Min Range: -0.5483

Max Act: 0.4449. Min Act: -0.0001

Data Index: 1302008 (C4 (Web Text))

Max Activating Token Index: 843

Text #6

Max Range: 0.5483. Min Range: -0.5483

Max Act: 0.4709. Min Act: -0.0000

Data Index: 939527 (C4 (Web Text))

Max Activating Token Index: 691

Text #7

Max Range: 0.5483. Min Range: -0.5483

Max Act: 0.4582. Min Act: -0.0001

Data Index: 438260 (C4 (Web Text))

Max Activating Token Index: 855

Text #8

Max Range: 0.5483. Min Range: -0.5483

Max Act: 0.4437. Min Act: -0.0001

Data Index: 785181 (C4 (Web Text))

Max Activating Token Index: 487

Text #9

Max Range: 0.5483. Min Range: -0.5483

Max Act: 0.3977. Min Act: -0.0001

Data Index: 64392 (C4 (Web Text))

Max Activating Token Index: 84

Text #10

Max Range: 0.5483. Min Range: -0.5483

Max Act: 0.4262. Min Act: -0.0001

Data Index: 742235 (C4 (Web Text))

Max Activating Token Index: 286

Text #11

Max Range: 0.5483. Min Range: -0.5483

Max Act: 0.4242. Min Act: -0.0001

Data Index: 1036219 (C4 (Web Text))

Max Activating Token Index: 722

Text #12

Max Range: 0.5483. Min Range: -0.5483

Max Act: 0.4176. Min Act: -0.0000

Data Index: 1310904 (C4 (Web Text))

Max Activating Token Index: 77

Text #13

Max Range: 0.5483. Min Range: -0.5483

Max Act: 0.3854. Min Act: -0.0000

Data Index: 1322408 (C4 (Web Text))

Max Activating Token Index: 141

Text #14

Max Range: 0.5483. Min Range: -0.5483

Max Act: 0.4003. Min Act: -0.0000

Data Index: 258914 (C4 (Web Text))

Max Activating Token Index: 117

Text #15

Max Range: 0.5483. Min Range: -0.5483

Max Act: 0.3946. Min Act: -0.0001

Data Index: 1380032 (Python Code)

Max Activating Token Index: 763

Text #16

Max Range: 0.5483. Min Range: -0.5483

Max Act: 0.4042. Min Act: -0.0001

Data Index: 1072425 (C4 (Web Text))

Max Activating Token Index: 794

Text #17

Max Range: 0.5483. Min Range: -0.5483

Max Act: 0.3876. Min Act: -0.0001

Data Index: 584773 (C4 (Web Text))

Max Activating Token Index: 896

Text #18

Max Range: 0.5483. Min Range: -0.5483

Max Act: 0.3610. Min Act: -0.0001

Data Index: 1193792 (C4 (Web Text))

Max Activating Token Index: 625

Text #19

Max Range: 0.5483. Min Range: -0.5483

Max Act: 0.3602. Min Act: -0.0001

Data Index: 932451 (C4 (Web Text))

Max Activating Token Index: 332