Model: SoLU Model: 6 Layers, 3072 Neurons per Layer

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

Neuron 2102 in Layer 4

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-6l')

Text #0

Max Range: 0.6174. Min Range: -0.6174

Max Act: 0.6174. Min Act: -0.0001

Data Index: 1054499 (C4 (Web Text))

Max Activating Token Index: 957

Text #1

Max Range: 0.6174. Min Range: -0.6174

Max Act: 0.5611. Min Act: -0.0001

Data Index: 1403239 (Python Code)

Max Activating Token Index: 220

Text #2

Max Range: 0.6174. Min Range: -0.6174

Max Act: 0.5784. Min Act: -0.0001

Data Index: 607161 (C4 (Web Text))

Max Activating Token Index: 700

Text #3

Max Range: 0.6174. Min Range: -0.6174

Max Act: 0.5934. Min Act: -0.0001

Data Index: 1324082 (C4 (Web Text))

Max Activating Token Index: 60

Text #4

Max Range: 0.6174. Min Range: -0.6174

Max Act: 0.5336. Min Act: -0.0001

Data Index: 66437 (C4 (Web Text))

Max Activating Token Index: 751

Text #5

Max Range: 0.6174. Min Range: -0.6174

Max Act: 0.5832. Min Act: -0.0001

Data Index: 540658 (C4 (Web Text))

Max Activating Token Index: 706

Text #6

Max Range: 0.6174. Min Range: -0.6174

Max Act: 0.5817. Min Act: -0.0001

Data Index: 882815 (C4 (Web Text))

Max Activating Token Index: 1004

Text #7

Max Range: 0.6174. Min Range: -0.6174

Max Act: 0.5591. Min Act: -0.0001

Data Index: 12997 (C4 (Web Text))

Max Activating Token Index: 487

Text #8

Max Range: 0.6174. Min Range: -0.6174

Max Act: 0.5474. Min Act: -0.0001

Data Index: 171608 (C4 (Web Text))

Max Activating Token Index: 105

Text #9

Max Range: 0.6174. Min Range: -0.6174

Max Act: 0.4903. Min Act: -0.0001

Data Index: 1246850 (C4 (Web Text))

Max Activating Token Index: 1009

Text #10

Max Range: 0.6174. Min Range: -0.6174

Max Act: 0.4972. Min Act: -0.0001

Data Index: 162290 (C4 (Web Text))

Max Activating Token Index: 140

Text #11

Max Range: 0.6174. Min Range: -0.6174

Max Act: 0.5090. Min Act: -0.0001

Data Index: 1047847 (C4 (Web Text))

Max Activating Token Index: 459

Text #12

Max Range: 0.6174. Min Range: -0.6174

Max Act: 0.4686. Min Act: -0.0001

Data Index: 839252 (C4 (Web Text))

Max Activating Token Index: 402

Text #13

Max Range: 0.6174. Min Range: -0.6174

Max Act: 0.4944. Min Act: -0.0001

Data Index: 823397 (C4 (Web Text))

Max Activating Token Index: 831

Text #14

Max Range: 0.6174. Min Range: -0.6174

Max Act: 0.4801. Min Act: -0.0001

Data Index: 750194 (C4 (Web Text))

Max Activating Token Index: 677

Text #15

Max Range: 0.6174. Min Range: -0.6174

Max Act: 0.4933. Min Act: -0.0001

Data Index: 655090 (C4 (Web Text))

Max Activating Token Index: 579

Text #16

Max Range: 0.6174. Min Range: -0.6174

Max Act: 0.5029. Min Act: -0.0001

Data Index: 592754 (C4 (Web Text))

Max Activating Token Index: 904

Text #17

Max Range: 0.6174. Min Range: -0.6174

Max Act: 0.4821. Min Act: -0.0001

Data Index: 402896 (C4 (Web Text))

Max Activating Token Index: 880

Text #18

Max Range: 0.6174. Min Range: -0.6174

Max Act: 0.4708. Min Act: -0.0001

Data Index: 10623 (C4 (Web Text))

Max Activating Token Index: 44

Text #19

Max Range: 0.6174. Min Range: -0.6174

Max Act: 0.4379. Min Act: -0.0001

Data Index: 454082 (C4 (Web Text))

Max Activating Token Index: 42