ConQUR: Mitigating Delusional Bias in Deep Q-Learning DiJia (Andy) - - PowerPoint PPT Presentation

▶

Dec 05, 2023 515 likes •858 views

ConQUR: Mitigating Delusional Bias in Deep Q-Learning DiJia (Andy) Su (Princeton) Jayden Ooi (Google) Tyler Lu (Google) Dale Schuurmans (Google) Craig Boutilier (Google) - Search:

SLIDE 1

DiJia (Andy) Su (Princeton) Jayden Ooi (Google) Tyler Lu (Google) Dale Schuurmans (Google) Craig Boutilier (Google)

ConQUR: Mitigating Delusional Bias in Deep Q-Learning

SLIDE 2

SLIDE 3

SLIDE 4

SLIDE 5

SLIDE 6

SLIDE 7

SLIDE 8

SLIDE 9

SLIDE 10

SLIDE 11

SLIDE 12

SLIDE 13

Search: Q-Regressors

P 13 Proprietary & Confidential

SLIDE 14

P 14 Proprietary & Confidential

Batch #1

………………………………………..
………………………………………..

SLIDE 15

P 15 Proprietary & Confidential

Batch #1

………………………………………..
………………………………………..

Batch #1

………………………………………..
………………………………………..

Generate & fit Q-labels G e n e r a t e & fi t Q

a b e l s

SLIDE 16

P 16 Proprietary & Confidential

Batch #1

………………………………………..
………………………………………..

Batch #1

………………………………………..
………………………………………..

Generate & fit Q-labels G e n e r a t e & fi t Q

a b e l s

SLIDE 17

P 17 Proprietary & Confidential

Batch #1

………………………………………..
………………………………………..

Batch #1

………………………………………..
………………………………………..

Generate & fit Q-labels G e n e r a t e & fi t Q

a b e l s

SLIDE 18

P 18 Proprietary & Confidential

Q-Label Set #1 Q-Label Set #2 Batch #1

………………………………………..
………………………………………..

Batch #1

………………………………………..
………………………………………..

Generate & fit Q-labels G e n e r a t e & fi t Q

a b e l s

SLIDE 19

P 19 Proprietary & Confidential

Batch #1

………………………………………..
………………………………………..

Batch #1

………………………………………..
………………………………………..

Q-Label Set #1 Q-Label Set #2

SLIDE 20

P 20 Proprietary & Confidential

Batch #1

………………………………………..
………………………………………..

Batch #1

………………………………………..
………………………………………..

Batch #2

………………………………………..
………………………………………..

SLIDE 21

P 21 Proprietary & Confidential

Batch #1

………………………………………..
………………………………………..

Batch #1

………………………………………..
………………………………………..

Batch #2

………………………………………..
………………………………………..

SLIDE 22

P 22 Proprietary & Confidential

Batch #1

………………………………………..
………………………………………..

Batch #1

………………………………………..
………………………………………..

Batch #2

………………………………………..
………………………………………..

SLIDE 23

P 23 Proprietary & Confidential

………

SLIDE 24

SLIDE 25

1)

SLIDE 26

SLIDE 27

SLIDE 28

SLIDE 29

SLIDE 30

SLIDE 31

SLIDE 32