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

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

Neuron 1922 in Layer 0

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 3.3731. Min Range: -3.3731

Max Act: 3.3731. Min Act: -0.1700

Data Index: 540445 (C4 (Web Text))

Max Activating Token Index: 599

Text #1

Max Range: 3.3731. Min Range: -3.3731

Max Act: 3.2789. Min Act: -0.1700

Data Index: 312972 (C4 (Web Text))

Max Activating Token Index: 766

Text #2

Max Range: 3.3731. Min Range: -3.3731

Max Act: 3.2474. Min Act: -0.1700

Data Index: 333345 (C4 (Web Text))

Max Activating Token Index: 773

Text #3

Max Range: 3.3731. Min Range: -3.3731

Max Act: 3.2317. Min Act: -0.1700

Data Index: 1172421 (C4 (Web Text))

Max Activating Token Index: 930

Text #4

Max Range: 3.3731. Min Range: -3.3731

Max Act: 3.2317. Min Act: -0.1700

Data Index: 467982 (C4 (Web Text))

Max Activating Token Index: 652

Text #5

Max Range: 3.3731. Min Range: -3.3731

Max Act: 3.2160. Min Act: -0.1700

Data Index: 1457637 (Python Code)

Max Activating Token Index: 633

Text #6

Max Range: 3.3731. Min Range: -3.3731

Max Act: 3.2002. Min Act: -0.1700

Data Index: 995467 (C4 (Web Text))

Max Activating Token Index: 677

Text #7

Max Range: 3.3731. Min Range: -3.3731

Max Act: 3.2002. Min Act: -0.1700

Data Index: 337426 (C4 (Web Text))

Max Activating Token Index: 316

Text #8

Max Range: 3.3731. Min Range: -3.3731

Max Act: 3.1688. Min Act: -0.1700

Data Index: 1171453 (C4 (Web Text))

Max Activating Token Index: 568

Text #9

Max Range: 3.3731. Min Range: -3.3731

Max Act: 3.1688. Min Act: -0.1700

Data Index: 522125 (C4 (Web Text))

Max Activating Token Index: 584

Text #10

Max Range: 3.3731. Min Range: -3.3731

Max Act: 3.1845. Min Act: -0.1700

Data Index: 22921 (C4 (Web Text))

Max Activating Token Index: 311

Text #11

Max Range: 3.3731. Min Range: -3.3731

Max Act: 3.1530. Min Act: -0.1700

Data Index: 1355716 (C4 (Web Text))

Max Activating Token Index: 227

Text #12

Max Range: 3.3731. Min Range: -3.3731

Max Act: 3.1373. Min Act: -0.1700

Data Index: 1303316 (C4 (Web Text))

Max Activating Token Index: 506

Text #13

Max Range: 3.3731. Min Range: -3.3731

Max Act: 3.1530. Min Act: -0.1700

Data Index: 167949 (C4 (Web Text))

Max Activating Token Index: 350

Text #14

Max Range: 3.3731. Min Range: -3.3731

Max Act: 3.1530. Min Act: -0.1700

Data Index: 312991 (C4 (Web Text))

Max Activating Token Index: 241

Text #15

Max Range: 3.3731. Min Range: -3.3731

Max Act: 3.1530. Min Act: -0.1700

Data Index: 1606457 (Python Code)

Max Activating Token Index: 764

Text #16

Max Range: 3.3731. Min Range: -3.3731

Max Act: 3.1373. Min Act: -0.1700

Data Index: 186238 (C4 (Web Text))

Max Activating Token Index: 526

Text #17

Max Range: 3.3731. Min Range: -3.3731

Max Act: 3.1215. Min Act: -0.1700

Data Index: 1272321 (C4 (Web Text))

Max Activating Token Index: 35

Text #18

Max Range: 3.3731. Min Range: -3.3731

Max Act: 3.1373. Min Act: -0.1700

Data Index: 660738 (C4 (Web Text))

Max Activating Token Index: 154

Text #19

Max Range: 3.3731. Min Range: -3.3731

Max Act: 3.1373. Min Act: -0.1700

Data Index: 1031884 (C4 (Web Text))

Max Activating Token Index: 1003