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

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

Neuron 863 in Layer 3

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 5.3864. Min Range: -5.3864

Max Act: 5.3864. Min Act: -0.1700

Data Index: 1076302 (C4 (Web Text))

Max Activating Token Index: 343

Text #1

Max Range: 5.3864. Min Range: -5.3864

Max Act: 5.2614. Min Act: -0.1700

Data Index: 990889 (C4 (Web Text))

Max Activating Token Index: 356

Text #2

Max Range: 5.3864. Min Range: -5.3864

Max Act: 5.3395. Min Act: -0.1700

Data Index: 1643968 (Python Code)

Max Activating Token Index: 632

Text #3

Max Range: 5.3864. Min Range: -5.3864

Max Act: 5.2145. Min Act: -0.1700

Data Index: 1429112 (Python Code)

Max Activating Token Index: 1001

Text #4

Max Range: 5.3864. Min Range: -5.3864

Max Act: 5.2145. Min Act: -0.1700

Data Index: 873149 (C4 (Web Text))

Max Activating Token Index: 382

Text #5

Max Range: 5.3864. Min Range: -5.3864

Max Act: 5.1833. Min Act: -0.1700

Data Index: 420126 (C4 (Web Text))

Max Activating Token Index: 420

Text #6

Max Range: 5.3864. Min Range: -5.3864

Max Act: 5.1676. Min Act: -0.1700

Data Index: 702865 (C4 (Web Text))

Max Activating Token Index: 555

Text #7

Max Range: 5.3864. Min Range: -5.3864

Max Act: 5.1989. Min Act: -0.1700

Data Index: 511486 (C4 (Web Text))

Max Activating Token Index: 543

Text #8

Max Range: 5.3864. Min Range: -5.3864

Max Act: 5.1520. Min Act: -0.1700

Data Index: 835120 (C4 (Web Text))

Max Activating Token Index: 508

Text #9

Max Range: 5.3864. Min Range: -5.3864

Max Act: 5.1051. Min Act: -0.1700

Data Index: 100228 (C4 (Web Text))

Max Activating Token Index: 809

Text #10

Max Range: 5.3864. Min Range: -5.3864

Max Act: 5.1364. Min Act: -0.1700

Data Index: 1107276 (C4 (Web Text))

Max Activating Token Index: 533

Text #11

Max Range: 5.3864. Min Range: -5.3864

Max Act: 5.1051. Min Act: -0.1700

Data Index: 1324266 (C4 (Web Text))

Max Activating Token Index: 221

Text #12

Max Range: 5.3864. Min Range: -5.3864

Max Act: 5.1676. Min Act: -0.1700

Data Index: 487671 (C4 (Web Text))

Max Activating Token Index: 513

Text #13

Max Range: 5.3864. Min Range: -5.3864

Max Act: 5.0739. Min Act: -0.1700

Data Index: 1263463 (C4 (Web Text))

Max Activating Token Index: 805

Text #14

Max Range: 5.3864. Min Range: -5.3864

Max Act: 5.1051. Min Act: -0.1700

Data Index: 536748 (C4 (Web Text))

Max Activating Token Index: 274

Text #15

Max Range: 5.3864. Min Range: -5.3864

Max Act: 5.0583. Min Act: -0.1700

Data Index: 206676 (C4 (Web Text))

Max Activating Token Index: 424

Text #16

Max Range: 5.3864. Min Range: -5.3864

Max Act: 5.0739. Min Act: -0.1700

Data Index: 202028 (C4 (Web Text))

Max Activating Token Index: 544

Text #17

Max Range: 5.3864. Min Range: -5.3864

Max Act: 5.0895. Min Act: -0.1700

Data Index: 146661 (C4 (Web Text))

Max Activating Token Index: 194

Text #18

Max Range: 5.3864. Min Range: -5.3864

Max Act: 5.1051. Min Act: -0.1700

Data Index: 955436 (C4 (Web Text))

Max Activating Token Index: 262

Text #19

Max Range: 5.3864. Min Range: -5.3864

Max Act: 5.1051. Min Act: -0.1699

Data Index: 675501 (C4 (Web Text))

Max Activating Token Index: 544