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

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

Neuron 250 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: 3.4618. Min Range: -3.4618

Max Act: 3.4618. Min Act: -0.1700

Data Index: 1422034 (Python Code)

Max Activating Token Index: 508

Text #1

Max Range: 3.4618. Min Range: -3.4618

Max Act: 3.1948. Min Act: -0.1700

Data Index: 1310686 (C4 (Web Text))

Max Activating Token Index: 316

Text #2

Max Range: 3.4618. Min Range: -3.4618

Max Act: 3.1948. Min Act: -0.1700

Data Index: 1274534 (C4 (Web Text))

Max Activating Token Index: 447

Text #3

Max Range: 3.4618. Min Range: -3.4618

Max Act: 3.1476. Min Act: -0.1700

Data Index: 932199 (C4 (Web Text))

Max Activating Token Index: 831

Text #4

Max Range: 3.4618. Min Range: -3.4618

Max Act: 3.1476. Min Act: -0.1700

Data Index: 1523740 (Python Code)

Max Activating Token Index: 501

Text #5

Max Range: 3.4618. Min Range: -3.4618

Max Act: 3.1319. Min Act: -0.1700

Data Index: 1111619 (C4 (Web Text))

Max Activating Token Index: 549

Text #6

Max Range: 3.4618. Min Range: -3.4618

Max Act: 3.1161. Min Act: -0.1700

Data Index: 212558 (C4 (Web Text))

Max Activating Token Index: 389

Text #7

Max Range: 3.4618. Min Range: -3.4618

Max Act: 3.1319. Min Act: -0.1700

Data Index: 633079 (C4 (Web Text))

Max Activating Token Index: 27

Text #8

Max Range: 3.4618. Min Range: -3.4618

Max Act: 3.0688. Min Act: -0.1700

Data Index: 692356 (C4 (Web Text))

Max Activating Token Index: 427

Text #9

Max Range: 3.4618. Min Range: -3.4618

Max Act: 3.0846. Min Act: -0.1700

Data Index: 309423 (C4 (Web Text))

Max Activating Token Index: 237

Text #10

Max Range: 3.4618. Min Range: -3.4618

Max Act: 3.0846. Min Act: -0.1700

Data Index: 103602 (C4 (Web Text))

Max Activating Token Index: 758

Text #11

Max Range: 3.4618. Min Range: -3.4618

Max Act: 3.0530. Min Act: -0.1700

Data Index: 991600 (C4 (Web Text))

Max Activating Token Index: 466

Text #12

Max Range: 3.4618. Min Range: -3.4618

Max Act: 3.0688. Min Act: -0.1700

Data Index: 1028927 (C4 (Web Text))

Max Activating Token Index: 655

Text #13

Max Range: 3.4618. Min Range: -3.4618

Max Act: 3.0688. Min Act: -0.1700

Data Index: 1477502 (Python Code)

Max Activating Token Index: 764

Text #14

Max Range: 3.4618. Min Range: -3.4618

Max Act: 3.0688. Min Act: -0.1700

Data Index: 1290677 (C4 (Web Text))

Max Activating Token Index: 451

Text #15

Max Range: 3.4618. Min Range: -3.4618

Max Act: 3.0846. Min Act: -0.1700

Data Index: 285693 (C4 (Web Text))

Max Activating Token Index: 680

Text #16

Max Range: 3.4618. Min Range: -3.4618

Max Act: 3.0372. Min Act: -0.1700

Data Index: 493171 (C4 (Web Text))

Max Activating Token Index: 373

Text #17

Max Range: 3.4618. Min Range: -3.4618

Max Act: 3.0530. Min Act: -0.1700

Data Index: 1341881 (C4 (Web Text))

Max Activating Token Index: 633

Text #18

Max Range: 3.4618. Min Range: -3.4618

Max Act: 3.0530. Min Act: -0.1700

Data Index: 364961 (C4 (Web Text))

Max Activating Token Index: 234

Text #19

Max Range: 3.4618. Min Range: -3.4618

Max Act: 3.0530. Min Act: -0.1700

Data Index: 496924 (C4 (Web Text))

Max Activating Token Index: 503