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

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

Neuron 233 in Layer 2

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 5.6028. Min Range: -5.6028

Max Act: 5.6028. Min Act: -0.1700

Data Index: 1064115 (C4 (Web Text))

Max Activating Token Index: 665

Text #1

Max Range: 5.6028. Min Range: -5.6028

Max Act: 5.5716. Min Act: -0.1700

Data Index: 1633019 (Python Code)

Max Activating Token Index: 883

Text #2

Max Range: 5.6028. Min Range: -5.6028

Max Act: 5.4466. Min Act: -0.1700

Data Index: 1443435 (Python Code)

Max Activating Token Index: 965

Text #3

Max Range: 5.6028. Min Range: -5.6028

Max Act: 5.3216. Min Act: -0.1700

Data Index: 1579869 (Python Code)

Max Activating Token Index: 803

Text #4

Max Range: 5.6028. Min Range: -5.6028

Max Act: 5.2591. Min Act: -0.1700

Data Index: 1372948 (Python Code)

Max Activating Token Index: 599

Text #5

Max Range: 5.6028. Min Range: -5.6028

Max Act: 5.2591. Min Act: -0.1700

Data Index: 1631824 (Python Code)

Max Activating Token Index: 553

Text #6

Max Range: 5.6028. Min Range: -5.6028

Max Act: 5.2278. Min Act: -0.1700

Data Index: 1564747 (Python Code)

Max Activating Token Index: 868

Text #7

Max Range: 5.6028. Min Range: -5.6028

Max Act: 5.2278. Min Act: -0.1700

Data Index: 1609376 (Python Code)

Max Activating Token Index: 339

Text #8

Max Range: 5.6028. Min Range: -5.6028

Max Act: 5.1653. Min Act: -0.1700

Data Index: 1278007 (C4 (Web Text))

Max Activating Token Index: 231

Text #9

Max Range: 5.6028. Min Range: -5.6028

Max Act: 5.1653. Min Act: -0.1700

Data Index: 1547468 (Python Code)

Max Activating Token Index: 846

Text #10

Max Range: 5.6028. Min Range: -5.6028

Max Act: 5.1185. Min Act: -0.1700

Data Index: 1652485 (Python Code)

Max Activating Token Index: 592

Text #11

Max Range: 5.6028. Min Range: -5.6028

Max Act: 5.1341. Min Act: -0.1700

Data Index: 1503586 (Python Code)

Max Activating Token Index: 1009

Text #12

Max Range: 5.6028. Min Range: -5.6028

Max Act: 5.0560. Min Act: -0.1700

Data Index: 1613626 (Python Code)

Max Activating Token Index: 877

Text #13

Max Range: 5.6028. Min Range: -5.6028

Max Act: 5.1185. Min Act: -0.1700

Data Index: 1376259 (Python Code)

Max Activating Token Index: 283

Text #14

Max Range: 5.6028. Min Range: -5.6028

Max Act: 5.0872. Min Act: -0.1700

Data Index: 1421286 (Python Code)

Max Activating Token Index: 404

Text #15

Max Range: 5.6028. Min Range: -5.6028

Max Act: 5.0716. Min Act: -0.1700

Data Index: 1557753 (Python Code)

Max Activating Token Index: 170

Text #16

Max Range: 5.6028. Min Range: -5.6028

Max Act: 5.0872. Min Act: -0.1700

Data Index: 1625403 (Python Code)

Max Activating Token Index: 384

Text #17

Max Range: 5.6028. Min Range: -5.6028

Max Act: 5.0560. Min Act: -0.1700

Data Index: 1589458 (Python Code)

Max Activating Token Index: 38

Text #18

Max Range: 5.6028. Min Range: -5.6028

Max Act: 4.9778. Min Act: -0.1700

Data Index: 1434198 (Python Code)

Max Activating Token Index: 379

Text #19

Max Range: 5.6028. Min Range: -5.6028

Max Act: 5.0560. Min Act: -0.1700

Data Index: 1051075 (C4 (Web Text))

Max Activating Token Index: 695