Model: SoLU Model: 12 Layers, 6144 Neurons per Layer

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

Neuron 2396 in Layer 4

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-12l')

Text #0

Max Range: 1.7814. Min Range: -1.7814

Max Act: 1.7814. Min Act: -0.0001

Data Index: 1239782 (C4 (Web Text))

Max Activating Token Index: 804

Text #1

Max Range: 1.7814. Min Range: -1.7814

Max Act: 1.6058. Min Act: -0.0001

Data Index: 124177 (C4 (Web Text))

Max Activating Token Index: 876

Text #2

Max Range: 1.7814. Min Range: -1.7814

Max Act: 1.6066. Min Act: -0.0001

Data Index: 718346 (C4 (Web Text))

Max Activating Token Index: 229

Text #3

Max Range: 1.7814. Min Range: -1.7814

Max Act: 1.4257. Min Act: -0.0001

Data Index: 242095 (C4 (Web Text))

Max Activating Token Index: 239

Text #4

Max Range: 1.7814. Min Range: -1.7814

Max Act: 1.4112. Min Act: -0.0001

Data Index: 634833 (C4 (Web Text))

Max Activating Token Index: 163

Text #5

Max Range: 1.7814. Min Range: -1.7814

Max Act: 1.3053. Min Act: -0.0001

Data Index: 712276 (C4 (Web Text))

Max Activating Token Index: 3

Text #6

Max Range: 1.7814. Min Range: -1.7814

Max Act: 1.2498. Min Act: -0.0001

Data Index: 179652 (C4 (Web Text))

Max Activating Token Index: 225

Text #7

Max Range: 1.7814. Min Range: -1.7814

Max Act: 1.2653. Min Act: -0.0001

Data Index: 198047 (C4 (Web Text))

Max Activating Token Index: 691

Text #8

Max Range: 1.7814. Min Range: -1.7814

Max Act: 1.2105. Min Act: -0.0001

Data Index: 244961 (C4 (Web Text))

Max Activating Token Index: 154

Text #9

Max Range: 1.7814. Min Range: -1.7814

Max Act: 1.2738. Min Act: -0.0001

Data Index: 477555 (C4 (Web Text))

Max Activating Token Index: 502

Text #10

Max Range: 1.7814. Min Range: -1.7814

Max Act: 1.2690. Min Act: -0.0001

Data Index: 102582 (C4 (Web Text))

Max Activating Token Index: 821

Text #11

Max Range: 1.7814. Min Range: -1.7814

Max Act: 1.3917. Min Act: -0.0001

Data Index: 1270550 (C4 (Web Text))

Max Activating Token Index: 882

Text #12

Max Range: 1.7814. Min Range: -1.7814

Max Act: 1.2586. Min Act: -0.0001

Data Index: 280200 (C4 (Web Text))

Max Activating Token Index: 867

Text #13

Max Range: 1.7814. Min Range: -1.7814

Max Act: 1.1707. Min Act: -0.0001

Data Index: 1470025 (Python Code)

Max Activating Token Index: 845

Text #14

Max Range: 1.7814. Min Range: -1.7814

Max Act: 1.2408. Min Act: -0.0001

Data Index: 73249 (C4 (Web Text))

Max Activating Token Index: 745

Text #15

Max Range: 1.7814. Min Range: -1.7814

Max Act: 1.2878. Min Act: -0.0001

Data Index: 1042240 (C4 (Web Text))

Max Activating Token Index: 50

Text #16

Max Range: 1.7814. Min Range: -1.7814

Max Act: 1.1926. Min Act: -0.0001

Data Index: 1510081 (Python Code)

Max Activating Token Index: 665

Text #17

Max Range: 1.7814. Min Range: -1.7814

Max Act: 1.3460. Min Act: -0.0001

Data Index: 1026557 (C4 (Web Text))

Max Activating Token Index: 946

Text #18

Max Range: 1.7814. Min Range: -1.7814

Max Act: 1.1757. Min Act: -0.0001

Data Index: 588420 (C4 (Web Text))

Max Activating Token Index: 695

Text #19

Max Range: 1.7814. Min Range: -1.7814

Max Act: 1.0964. Min Act: -0.0001

Data Index: 1578500 (Python Code)

Max Activating Token Index: 770