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

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

Neuron 2312 in Layer 6

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 1.8621. Min Range: -1.8621

Max Act: 1.8621. Min Act: -0.0001

Data Index: 1375681 (Python Code)

Max Activating Token Index: 397

Text #1

Max Range: 1.8621. Min Range: -1.8621

Max Act: 1.6009. Min Act: -0.0001

Data Index: 737844 (C4 (Web Text))

Max Activating Token Index: 484

Text #2

Max Range: 1.8621. Min Range: -1.8621

Max Act: 1.5584. Min Act: -0.0000

Data Index: 959409 (C4 (Web Text))

Max Activating Token Index: 271

Text #3

Max Range: 1.8621. Min Range: -1.8621

Max Act: 1.4783. Min Act: -0.0001

Data Index: 71855 (C4 (Web Text))

Max Activating Token Index: 265

Text #4

Max Range: 1.8621. Min Range: -1.8621

Max Act: 1.4685. Min Act: -0.0000

Data Index: 1147160 (C4 (Web Text))

Max Activating Token Index: 982

Text #5

Max Range: 1.8621. Min Range: -1.8621

Max Act: 1.4057. Min Act: -0.0000

Data Index: 76022 (C4 (Web Text))

Max Activating Token Index: 897

Text #6

Max Range: 1.8621. Min Range: -1.8621

Max Act: 1.4887. Min Act: -0.0001

Data Index: 1512004 (Python Code)

Max Activating Token Index: 577

Text #7

Max Range: 1.8621. Min Range: -1.8621

Max Act: 1.5250. Min Act: -0.0001

Data Index: 6368 (C4 (Web Text))

Max Activating Token Index: 510

Text #8

Max Range: 1.8621. Min Range: -1.8621

Max Act: 1.4728. Min Act: -0.0001

Data Index: 1498717 (Python Code)

Max Activating Token Index: 17

Text #9

Max Range: 1.8621. Min Range: -1.8621

Max Act: 1.3787. Min Act: -0.0000

Data Index: 30272 (C4 (Web Text))

Max Activating Token Index: 151

Text #10

Max Range: 1.8621. Min Range: -1.8621

Max Act: 1.2935. Min Act: -0.0001

Data Index: 1598301 (Python Code)

Max Activating Token Index: 113

Text #11

Max Range: 1.8621. Min Range: -1.8621

Max Act: 1.3545. Min Act: -0.0000

Data Index: 1140805 (C4 (Web Text))

Max Activating Token Index: 543

Text #12

Max Range: 1.8621. Min Range: -1.8621

Max Act: 1.3015. Min Act: -0.0001

Data Index: 1520633 (Python Code)

Max Activating Token Index: 501

Text #13

Max Range: 1.8621. Min Range: -1.8621

Max Act: 1.4585. Min Act: -0.0001

Data Index: 1508661 (Python Code)

Max Activating Token Index: 646

Text #14

Max Range: 1.8621. Min Range: -1.8621

Max Act: 1.3265. Min Act: -0.0000

Data Index: 1312296 (C4 (Web Text))

Max Activating Token Index: 95

Text #15

Max Range: 1.8621. Min Range: -1.8621

Max Act: 1.2422. Min Act: -0.0001

Data Index: 1052948 (C4 (Web Text))

Max Activating Token Index: 290

Text #16

Max Range: 1.8621. Min Range: -1.8621

Max Act: 1.3182. Min Act: -0.0001

Data Index: 1638238 (Python Code)

Max Activating Token Index: 911

Text #17

Max Range: 1.8621. Min Range: -1.8621

Max Act: 1.2557. Min Act: -0.0000

Data Index: 113855 (C4 (Web Text))

Max Activating Token Index: 978

Text #18

Max Range: 1.8621. Min Range: -1.8621

Max Act: 1.2911. Min Act: -0.0000

Data Index: 1189294 (C4 (Web Text))

Max Activating Token Index: 754

Text #19

Max Range: 1.8621. Min Range: -1.8621

Max Act: 1.2775. Min Act: -0.0000

Data Index: 361202 (C4 (Web Text))

Max Activating Token Index: 537