Model: SoLU Model: 3 Layers, 2048 Neurons per Layer

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

Neuron 1380 in Layer 2

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-3l')

Text #0

Max Range: 0.5740. Min Range: -0.5740

Max Act: 0.5740. Min Act: -0.0003

Data Index: 704806 (C4 (Web Text))

Max Activating Token Index: 756

Text #1

Max Range: 0.5740. Min Range: -0.5740

Max Act: 0.4165. Min Act: -0.0003

Data Index: 1054689 (C4 (Web Text))

Max Activating Token Index: 89

Text #2

Max Range: 0.5740. Min Range: -0.5740

Max Act: 0.4057. Min Act: -0.0003

Data Index: 356432 (C4 (Web Text))

Max Activating Token Index: 789

Text #3

Max Range: 0.5740. Min Range: -0.5740

Max Act: 0.3866. Min Act: -0.0003

Data Index: 1233905 (C4 (Web Text))

Max Activating Token Index: 758

Text #4

Max Range: 0.5740. Min Range: -0.5740

Max Act: 0.3527. Min Act: -0.0003

Data Index: 847443 (C4 (Web Text))

Max Activating Token Index: 318

Text #5

Max Range: 0.5740. Min Range: -0.5740

Max Act: 0.3461. Min Act: -0.0003

Data Index: 938575 (C4 (Web Text))

Max Activating Token Index: 723

Text #6

Max Range: 0.5740. Min Range: -0.5740

Max Act: 0.3145. Min Act: -0.0003

Data Index: 990267 (C4 (Web Text))

Max Activating Token Index: 421

Text #7

Max Range: 0.5740. Min Range: -0.5740

Max Act: 0.3267. Min Act: -0.0003

Data Index: 141185 (C4 (Web Text))

Max Activating Token Index: 581

Text #8

Max Range: 0.5740. Min Range: -0.5740

Max Act: 0.3276. Min Act: -0.0003

Data Index: 55055 (C4 (Web Text))

Max Activating Token Index: 554

Text #9

Max Range: 0.5740. Min Range: -0.5740

Max Act: 0.3387. Min Act: -0.0003

Data Index: 496250 (C4 (Web Text))

Max Activating Token Index: 652

Text #10

Max Range: 0.5740. Min Range: -0.5740

Max Act: 0.3267. Min Act: -0.0003

Data Index: 510233 (C4 (Web Text))

Max Activating Token Index: 676

Text #11

Max Range: 0.5740. Min Range: -0.5740

Max Act: 0.3152. Min Act: -0.0003

Data Index: 857250 (C4 (Web Text))

Max Activating Token Index: 700

Text #12

Max Range: 0.5740. Min Range: -0.5740

Max Act: 0.3111. Min Act: -0.0003

Data Index: 441578 (C4 (Web Text))

Max Activating Token Index: 841

Text #13

Max Range: 0.5740. Min Range: -0.5740

Max Act: 0.2954. Min Act: -0.0003

Data Index: 218729 (C4 (Web Text))

Max Activating Token Index: 568

Text #14

Max Range: 0.5740. Min Range: -0.5740

Max Act: 0.3158. Min Act: -0.0003

Data Index: 922412 (C4 (Web Text))

Max Activating Token Index: 701

Text #15

Max Range: 0.5740. Min Range: -0.5740

Max Act: 0.3070. Min Act: -0.0003

Data Index: 530681 (C4 (Web Text))

Max Activating Token Index: 794

Text #16

Max Range: 0.5740. Min Range: -0.5740

Max Act: 0.3045. Min Act: -0.0003

Data Index: 325321 (C4 (Web Text))

Max Activating Token Index: 541

Text #17

Max Range: 0.5740. Min Range: -0.5740

Max Act: 0.3018. Min Act: -0.0003

Data Index: 1113645 (C4 (Web Text))

Max Activating Token Index: 271

Text #18

Max Range: 0.5740. Min Range: -0.5740

Max Act: 0.3002. Min Act: -0.0003

Data Index: 263195 (C4 (Web Text))

Max Activating Token Index: 542

Text #19

Max Range: 0.5740. Min Range: -0.5740

Max Act: 0.2846. Min Act: -0.0003

Data Index: 861349 (C4 (Web Text))

Max Activating Token Index: 794