I’ll retweet this every time i see it

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

1/ Some initial thoughts on personal moats:

Like company moats, your personal moat should be a competitive advantage that is not only durable—it should also compound over time.

Characteristics of a personal moat below:


2/ Like a company moat, you want to build career capital while you sleep.

As Andrew Chen noted:


3/ You don’t want to build a competitive advantage that is fleeting or that will get commoditized

Things that might get commoditized over time (some longer than


4/ Before the arrival of recorded music, what used to be scarce was the actual music itself — required an in-person artist.

After recorded music, the music itself became abundant and what became scarce was curation, distribution, and self space.

5/ Similarly, in careers, what used to be (more) scarce were things like ideas, money, and exclusive relationships.

In the internet economy, what has become scarce are things like specific knowledge, rare & valuable skills, and great reputations.
Still wondering about this 🤔


save as q