Episode 16: Yilun Du, MIT, on energy-based models, implicit functions, and modularity

1

Introduction

00:00 • 2min

2

How Have Your Interests Evolved Over Time?

01:56 • 2min

3

Is There a Change in the Ability of Things You Can Do With Them?

03:33 • 2min

4

Is It Possible to Create a Very Intelligent Agent Using Text?

05:51 • 2min

5

The Evolution of Multi Agent Systems

08:12 • 2min

6

Is There a Limiting Factor?

10:09 • 2min

7

Is Reactor Learning the Best Way to Train Your Agents?

12:02 • 4min

8

Is Your Policy Always Improved?

15:44 • 2min

9

Is Pure Randomness the Same Thing or Different?

17:23 • 2min

10

Learning Could See by Looking Boys, Um, Cool.

19:26 • 5min

11

How Do You Like Skill This Curiosity?

24:23 • 3min

12

How Did You Start Working on Your 20 19 Paper?

27:02 • 4min

13

How to Compose Unsupervised Energy Functions

30:46 • 3min

14

How Much Did You Play With That Reconstruction?

33:36 • 2min

15

Is It a Collusion or a Damage Enclusion?

35:12 • 2min

16

Do You Have a Ton of Energy Functions?

37:00 • 2min

17

I'm Not Sure Why It Gets Disentangled.

38:42 • 3min

18

Is There a Future Work at Ning Be Excited About the Paper?

41:25 • 2min

19

Are You Trying to Beat the State of the Art?

43:27 • 2min

20

The Benchmarking of Machine Learning

45:56 • 2min

21

The Idea of Modularity in Generative Modeling

47:53 • 2min

22

Is Modularity Important for Deep Learning?

49:24 • 2min

23

Is There Anyone Else Whose Work You Like in This Area?

51:19 • 2min

24

The Robotic Arm, Is It Really a Robotic Butler?

52:59 • 2min

25

Is It Possible to Use Modularity in Robotics?

54:47 • 2min

26

Using Neural Radiances to Represent Different Modalities?

57:03 • 2min

27

The Multimodality of a Modeling Approach?

59:02 • 3min

28

Dolty, What's the Most Interesting Thing About This Approach?

01:01:47 • 2min

29

Is There a Difference Between Functions and Structured Mesh Base Representations?

01:04:09 • 2min

30

Yer, More Ops?

01:06:19 • 2min

31

How to Use Energy Functions to Reasoning

01:08:33 • 2min

32

The Attractor Dynamic Theory of Memory

01:10:41 • 3min

33

Is There Anything You're Finding Very Interesting?

01:13:22 • 2min

34

Are There Any Arians Wha, You're Really Excited to See Them Develop?

01:14:58 • 3min

35

Are There Any Mistakes You've Made?

01:18:23 • 2min

36

I'm Not Sure if It's Better Than the Sateof the Arden?

01:20:10 • 2min

37

The New Comfen's Idea, No Answerable Ide

01:22:39 • 3min