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

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

Neuron 2704 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: 1.2609. Min Range: -1.2609

Max Act: 1.2609. Min Act: -0.0001

Data Index: 744577 (C4 (Web Text))

Max Activating Token Index: 837

Text #1

Max Range: 1.2609. Min Range: -1.2609

Max Act: 1.1631. Min Act: -0.0001

Data Index: 101577 (C4 (Web Text))

Max Activating Token Index: 889

Text #2

Max Range: 1.2609. Min Range: -1.2609

Max Act: 1.0807. Min Act: -0.0001

Data Index: 596155 (C4 (Web Text))

Max Activating Token Index: 613

Text #3

Max Range: 1.2609. Min Range: -1.2609

Max Act: 1.1069. Min Act: -0.0001

Data Index: 688305 (C4 (Web Text))

Max Activating Token Index: 944

Text #4

Max Range: 1.2609. Min Range: -1.2609

Max Act: 1.0454. Min Act: -0.0001

Data Index: 1095829 (C4 (Web Text))

Max Activating Token Index: 904

Text #5

Max Range: 1.2609. Min Range: -1.2609

Max Act: 1.1109. Min Act: -0.0001

Data Index: 1647099 (Python Code)

Max Activating Token Index: 425

Text #6

Max Range: 1.2609. Min Range: -1.2609

Max Act: 1.0908. Min Act: -0.0001

Data Index: 106094 (C4 (Web Text))

Max Activating Token Index: 640

Text #7

Max Range: 1.2609. Min Range: -1.2609

Max Act: 1.1016. Min Act: -0.0001

Data Index: 424076 (C4 (Web Text))

Max Activating Token Index: 192

Text #8

Max Range: 1.2609. Min Range: -1.2609

Max Act: 1.1012. Min Act: -0.0001

Data Index: 368070 (C4 (Web Text))

Max Activating Token Index: 348

Text #9

Max Range: 1.2609. Min Range: -1.2609

Max Act: 1.0867. Min Act: -0.0001

Data Index: 1183541 (C4 (Web Text))

Max Activating Token Index: 630

Text #10

Max Range: 1.2609. Min Range: -1.2609

Max Act: 1.0451. Min Act: -0.0001

Data Index: 642832 (C4 (Web Text))

Max Activating Token Index: 124

Text #11

Max Range: 1.2609. Min Range: -1.2609

Max Act: 0.9489. Min Act: -0.0001

Data Index: 200579 (C4 (Web Text))

Max Activating Token Index: 265

Text #12

Max Range: 1.2609. Min Range: -1.2609

Max Act: 1.0345. Min Act: -0.0001

Data Index: 1396190 (Python Code)

Max Activating Token Index: 418

Text #13

Max Range: 1.2609. Min Range: -1.2609

Max Act: 1.0195. Min Act: -0.0001

Data Index: 1105031 (C4 (Web Text))

Max Activating Token Index: 607

Text #14

Max Range: 1.2609. Min Range: -1.2609

Max Act: 0.9710. Min Act: -0.0001

Data Index: 747059 (C4 (Web Text))

Max Activating Token Index: 944

Text #15

Max Range: 1.2609. Min Range: -1.2609

Max Act: 0.9645. Min Act: -0.0001

Data Index: 1356396 (C4 (Web Text))

Max Activating Token Index: 228

Text #16

Max Range: 1.2609. Min Range: -1.2609

Max Act: 1.0239. Min Act: -0.0001

Data Index: 1265895 (C4 (Web Text))

Max Activating Token Index: 174

Text #17

Max Range: 1.2609. Min Range: -1.2609

Max Act: 1.0088. Min Act: -0.0001

Data Index: 792455 (C4 (Web Text))

Max Activating Token Index: 401

Text #18

Max Range: 1.2609. Min Range: -1.2609

Max Act: 0.9160. Min Act: -0.0001

Data Index: 495835 (C4 (Web Text))

Max Activating Token Index: 483

Text #19

Max Range: 1.2609. Min Range: -1.2609

Max Act: 0.9986. Min Act: -0.0001

Data Index: 493658 (C4 (Web Text))

Max Activating Token Index: 411