Model: SoLU Model: 4 Layers, 2048 Neurons per Layer

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

Neuron 128 in Layer 3

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-4l')

Text #0

Max Range: 0.0465. Min Range: -0.0465

Max Act: 0.0465. Min Act: -0.0002

Data Index: 690445 (C4 (Web Text))

Max Activating Token Index: 499

Text #1

Max Range: 0.0465. Min Range: -0.0465

Max Act: 0.0455. Min Act: -0.0002

Data Index: 881239 (C4 (Web Text))

Max Activating Token Index: 67

Text #2

Max Range: 0.0465. Min Range: -0.0465

Max Act: 0.0446. Min Act: -0.0002

Data Index: 618417 (C4 (Web Text))

Max Activating Token Index: 47

Text #3

Max Range: 0.0465. Min Range: -0.0465

Max Act: 0.0446. Min Act: -0.0002

Data Index: 419563 (C4 (Web Text))

Max Activating Token Index: 799

Text #4

Max Range: 0.0465. Min Range: -0.0465

Max Act: 0.0443. Min Act: -0.0002

Data Index: 679351 (C4 (Web Text))

Max Activating Token Index: 171

Text #5

Max Range: 0.0465. Min Range: -0.0465

Max Act: 0.0449. Min Act: -0.0002

Data Index: 498460 (C4 (Web Text))

Max Activating Token Index: 936

Text #6

Max Range: 0.0465. Min Range: -0.0465

Max Act: 0.0447. Min Act: -0.0001

Data Index: 40073 (C4 (Web Text))

Max Activating Token Index: 380

Text #7

Max Range: 0.0465. Min Range: -0.0465

Max Act: 0.0422. Min Act: -0.0002

Data Index: 607983 (C4 (Web Text))

Max Activating Token Index: 576

Text #8

Max Range: 0.0465. Min Range: -0.0465

Max Act: 0.0428. Min Act: -0.0002

Data Index: 947861 (C4 (Web Text))

Max Activating Token Index: 452

Text #9

Max Range: 0.0465. Min Range: -0.0465

Max Act: 0.0424. Min Act: -0.0002

Data Index: 863216 (C4 (Web Text))

Max Activating Token Index: 715

Text #10

Max Range: 0.0465. Min Range: -0.0465

Max Act: 0.0432. Min Act: -0.0002

Data Index: 709738 (C4 (Web Text))

Max Activating Token Index: 14

Text #11

Max Range: 0.0465. Min Range: -0.0465

Max Act: 0.0418. Min Act: -0.0001

Data Index: 272034 (C4 (Web Text))

Max Activating Token Index: 251

Text #12

Max Range: 0.0465. Min Range: -0.0465

Max Act: 0.0416. Min Act: -0.0002

Data Index: 105061 (C4 (Web Text))

Max Activating Token Index: 505

Text #13

Max Range: 0.0465. Min Range: -0.0465

Max Act: 0.0410. Min Act: -0.0001

Data Index: 1000912 (C4 (Web Text))

Max Activating Token Index: 229

Text #14

Max Range: 0.0465. Min Range: -0.0465

Max Act: 0.0406. Min Act: -0.0002

Data Index: 446627 (C4 (Web Text))

Max Activating Token Index: 443

Text #15

Max Range: 0.0465. Min Range: -0.0465

Max Act: 0.0407. Min Act: -0.0002

Data Index: 510085 (C4 (Web Text))

Max Activating Token Index: 351

Text #16

Max Range: 0.0465. Min Range: -0.0465

Max Act: 0.0414. Min Act: -0.0002

Data Index: 504034 (C4 (Web Text))

Max Activating Token Index: 471

Text #17

Max Range: 0.0465. Min Range: -0.0465

Max Act: 0.0405. Min Act: -0.0002

Data Index: 248350 (C4 (Web Text))

Max Activating Token Index: 513

Text #18

Max Range: 0.0465. Min Range: -0.0465

Max Act: 0.0405. Min Act: -0.0002

Data Index: 131375 (C4 (Web Text))

Max Activating Token Index: 379

Text #19

Max Range: 0.0465. Min Range: -0.0465

Max Act: 0.0401. Min Act: -0.0002

Data Index: 1295919 (C4 (Web Text))

Max Activating Token Index: 228