Adversarial Machine Learning Research with Florian Tramèr

1

Introduction

00:00 • 2min

2

Announcing the Sponsor of the ZK Podcast

01:46 • 5min

3

The Side Channel Attacks and Privacy Tokens

06:42 • 5min

4

What's the Problem With Side Channel Attacks in Cryptocurrency?

11:21 • 6min

5

The Ronan Bridge Hack - Is the IP Address Encrypted?

16:56 • 4min

6

Are You a PhD Candidate at Google Research?

20:31 • 3min

7

How Machine Learning Training Works

23:53 • 5min

8

Is There a Breaking Point in Machine Learning?

28:26 • 3min

9

Is There a Theorem to Encrypt Machine Learning Processes?

31:19 • 4min

10

Is ZKML Predefined or Model Predefined?

35:48 • 2min

11

Is There a Security Break in ML?

37:38 • 2min

12

Attacks on the Training Data in Machine Learning

39:26 • 3min

13

Are You Trying to Break Machine Learning Models?

42:24 • 4min

14

Is There a Trade-Off Between Fairness and Privacy?

45:58 • 5min

15

Is There a Trade-Off Between Privacy and Robustness?

50:38 • 4min

16

Is Stable Diffusion a Better Model Than Dolly?

55:04 • 3min

17

Can Peer Review Be Fixed by Using More Anomaly Motivated Models?

58:22 • 2min

18

Does It Make for More Striking Results?

59:55 • 4min

19

Impactful Science

01:03:58 • 3min