LESSWRONGTags
LW

SERI MATS

EditHistorySubscribe

Help improve this page

EditHistorySubscribe

Help improve this page

Contributors

The Stanford Existential Risks Initiative ML Alignment Theory Scholars program.

https://www.serimats.org/

Posts tagged SERI MATS

Most Relevant

7

71SERI MATS Program - Winter 2022 CohortΩ

Ryan Kidd, Victor Warlop, Christian Smith

4mo

12

6

21Project proposal: Testing the IBP definition of agent

Jeremy Gillen, Thomas Larsen, JamesH

6mo

4

4

117Taking the parameters which seem to matter and rotating them until they don't

5mo

48

3

19Normative vs Descriptive Models of AgencyΩ

2d

0

2

91Soft optimization makes the value target biggerΩ

1mo

18

2

72Neural Tangent Kernel Distillation

Thomas Larsen, Jeremy Gillen

4mo

20

2

48Race Along Rashomon RidgeΩ

Stephen Fowler, Peter S. Park, MichaelEinhorn

7mo

15

2

47Conditioning Generative Models for AlignmentΩ

7mo

8

2

47Consequentialists: One-Way Pattern Traps

19d

3

2

40More findings on Memorization and double descentΩ

Marius Hobbhahn

3d

0

2

29Auditing games for high-level interpretabilityΩ

3mo

1

2

29Broad Basins and Data Compression

Jeremy Gillen, Stephen Fowler, Thomas Larsen

6mo

6

2

26The Ground Truth Problem (Or, Why Evaluating Interpretability Methods Is Hard)

3mo

2

2

20More findings on maximal data dimensionΩ

Marius Hobbhahn

2d

0

2

17Swap and Scale

5mo

3