BREAKING: SEC says Sen. Richard Burr had material nonpublic info re COVID economic impact.

After Burr dumped stock, he called his brother-in-law.

His brother-in-law called his stock broker **the next minute**

ProPublica had revealed last year that Burr and his brother-in-law, Gerald Fauth, who sits on the National Mediation Board, dumped stock the same day.

That they talked that day, and that Fauth called his broker one minute after their call, is new.

https://t.co/1kY1oQHEn4
Also new: the SEC alleges that Burr had material nonpublic information based on his committee roles and **through former staffers who were directing key aspects of the government response to the virus**

https://t.co/EiXjZ4oPS4
The SEC subpoenaed Fauth soon after ProPublica's story last year. They say he has stonewalled.

An SEC insider trading investigation into Burr and Fauth is ongoing.

https://t.co/EiXjZ4oPS4
When we broke this story last year, Burr's attorney told us Burr "did not coordinate his decision to trade on Feb. 13 with Mr. Fauth.”

Now we know Burr called Fauth after the trades, and Fauth called his broker the very next minute.

What did Burr say?

https://t.co/EiXjZ4oPS4

More from All

How can we use language supervision to learn better visual representations for robotics?

Introducing Voltron: Language-Driven Representation Learning for Robotics!

Paper: https://t.co/gIsRPtSjKz
Models: https://t.co/NOB3cpATYG
Evaluation: https://t.co/aOzQu95J8z

🧵👇(1 / 12)


Videos of humans performing everyday tasks (Something-Something-v2, Ego4D) offer a rich and diverse resource for learning representations for robotic manipulation.

Yet, an underused part of these datasets are the rich, natural language annotations accompanying each video. (2/12)

The Voltron framework offers a simple way to use language supervision to shape representation learning, building off of prior work in representations for robotics like MVP (
https://t.co/Pb0mk9hb4i) and R3M (https://t.co/o2Fkc3fP0e).

The secret is *balance* (3/12)

Starting with a masked autoencoder over frames from these video clips, make a choice:

1) Condition on language and improve our ability to reconstruct the scene.

2) Generate language given the visual representation and improve our ability to describe what's happening. (4/12)

By trading off *conditioning* and *generation* we show that we can learn 1) better representations than prior methods, and 2) explicitly shape the balance of low and high-level features captured.

Why is the ability to shape this balance important? (5/12)

You May Also Like