Model: GELU Model: 2 Layers, 2048 Neurons per Layer

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

Neuron 272 in Layer 1

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('gelu-2l')

Text #0

Max Range: 4.0683. Min Range: -4.0683

Max Act: 4.0683. Min Act: -0.1700

Data Index: 755658 (C4 (Web Text))

Max Activating Token Index: 576

Text #1

Max Range: 4.0683. Min Range: -4.0683

Max Act: 3.9120. Min Act: -0.1700

Data Index: 1035967 (C4 (Web Text))

Max Activating Token Index: 458

Text #2

Max Range: 4.0683. Min Range: -4.0683

Max Act: 3.9432. Min Act: -0.1700

Data Index: 396900 (C4 (Web Text))

Max Activating Token Index: 444

Text #3

Max Range: 4.0683. Min Range: -4.0683

Max Act: 3.9120. Min Act: -0.1700

Data Index: 1191769 (C4 (Web Text))

Max Activating Token Index: 153

Text #4

Max Range: 4.0683. Min Range: -4.0683

Max Act: 3.8807. Min Act: -0.1700

Data Index: 1025578 (C4 (Web Text))

Max Activating Token Index: 914

Text #5

Max Range: 4.0683. Min Range: -4.0683

Max Act: 3.8494. Min Act: -0.1700

Data Index: 515431 (C4 (Web Text))

Max Activating Token Index: 594

Text #6

Max Range: 4.0683. Min Range: -4.0683

Max Act: 3.8807. Min Act: -0.1700

Data Index: 265892 (C4 (Web Text))

Max Activating Token Index: 274

Text #7

Max Range: 4.0683. Min Range: -4.0683

Max Act: 3.8807. Min Act: -0.1700

Data Index: 792168 (C4 (Web Text))

Max Activating Token Index: 361

Text #8

Max Range: 4.0683. Min Range: -4.0683

Max Act: 3.8494. Min Act: -0.1700

Data Index: 503917 (C4 (Web Text))

Max Activating Token Index: 243

Text #9

Max Range: 4.0683. Min Range: -4.0683

Max Act: 3.8181. Min Act: -0.1700

Data Index: 1520533 (Python Code)

Max Activating Token Index: 552

Text #10

Max Range: 4.0683. Min Range: -4.0683

Max Act: 3.8025. Min Act: -0.1700

Data Index: 635394 (C4 (Web Text))

Max Activating Token Index: 551

Text #11

Max Range: 4.0683. Min Range: -4.0683

Max Act: 3.7869. Min Act: -0.1700

Data Index: 1562667 (Python Code)

Max Activating Token Index: 407

Text #12

Max Range: 4.0683. Min Range: -4.0683

Max Act: 3.7712. Min Act: -0.1700

Data Index: 57403 (C4 (Web Text))

Max Activating Token Index: 219

Text #13

Max Range: 4.0683. Min Range: -4.0683

Max Act: 3.7712. Min Act: -0.1700

Data Index: 888584 (C4 (Web Text))

Max Activating Token Index: 401

Text #14

Max Range: 4.0683. Min Range: -4.0683

Max Act: 3.7869. Min Act: -0.1700

Data Index: 671200 (C4 (Web Text))

Max Activating Token Index: 721

Text #15

Max Range: 4.0683. Min Range: -4.0683

Max Act: 3.8025. Min Act: -0.1700

Data Index: 1606018 (Python Code)

Max Activating Token Index: 299

Text #16

Max Range: 4.0683. Min Range: -4.0683

Max Act: 3.8181. Min Act: -0.1700

Data Index: 1356816 (C4 (Web Text))

Max Activating Token Index: 701

Text #17

Max Range: 4.0683. Min Range: -4.0683

Max Act: 3.7243. Min Act: -0.1700

Data Index: 46998 (C4 (Web Text))

Max Activating Token Index: 421

Text #18

Max Range: 4.0683. Min Range: -4.0683

Max Act: 3.8025. Min Act: -0.1700

Data Index: 1118071 (C4 (Web Text))

Max Activating Token Index: 921

Text #19

Max Range: 4.0683. Min Range: -4.0683

Max Act: 3.7712. Min Act: -0.1700

Data Index: 1011506 (C4 (Web Text))

Max Activating Token Index: 819