Justin Bayer
@usuallyuseless
Followers
713
Following
3K
Media
33
Statuses
2K
Student, scholar, teacher & builder of learning autonomous systems.
Munich
Joined December 2008
World models of scenes (incl. dynamics) allow both SLAM and prediction for model-based control. Let me show you our recent work (ICLR 2021) where we demonstrate that with realistic drone data. Here is a video with ground truth, filtered locations and long-step predictions.
2
4
23
RT @dpkingma: It's already the case that people's free will gets hijacked by screens for hours a day, with lots of negative consequences. A….
0
35
0
RT @SZ_TopNews: Vermüllte Zimmer, abgefackelte Möbel, Fäkalien überall: Die Knorrhütte an der Zugspitze wurde im Winter von Unbekannten ver….
sueddeutsche.de
Die Knorrhütte an der Zugspitze wurde im Winter von Unbekannten verwüstet. Und sie ist kein Einzelfall. Können sich die Menschen denn nicht mal in den Bergen benehmen?
0
2
0
RT @pfau: The war on science in the US is already affecting private sector research like AlphaFold. Bears repeating but the private sector….
0
108
0
RT @NandoDF: RL is not all you need, nor attention nor Bayesianism nor free energy minimisation, nor an age of first person experience. Suc….
0
195
0
RT @tomonezero: Another flaw with Claude Code: it tends to change the implementation specifically to pass unit tests after a few unsuccessf….
0
1
0
RT @togelius: Who do you trust most on matters of AI?.Someone who has worked on AI for 2 or 20 years? .Someone who works in academia or a l….
0
11
0
RT @usuallyuseless: Bayesplaining: take a well established method, express it as a series of crude approximations to a Bayesian approach, t….
0
66
0
RT @togelius: Today is the official release day for my little book on Artificial General Intelligence, published by MIT Press. It's availab….
0
165
0
Come do a Phd with me and my colleagues!
jobs.volkswagen-group.com
Doktorandin / Doktorand Machine Learning Research Lab (w/m/d)
1
2
6
in fact we even have invariance to position here 🤪.
@Alonso_GD Is F=ma knowledge? If yes, what is its position? if no, what is knowledge?.
0
0
1
RT @pfau: My basic political philosophy is that monopolies are bad but the decline of Bell Labs set back human civilization by several deca….
0
182
0
pretty good life advice from my guru @togelius (= msc advisor): accept you suck at most things and make that work for you.
I'm the first to admit that I suck at most things. But I've been pretty good at making that work for myself. For example, I have huge problems reading badly written text, so I learned to write readable text. I also don't like competing, so I find new areas or topics to work on,.
0
0
2
RT @NurderK: „Wir müssen raus aus diesem Trübsinn. Wir müssen den Rücken gerade machen und sagen: Da ist ein Problem, da sind zehn Probleme….
0
657
0
👇.
Identifying the true structure of the latent data generating process (causal world model) by learning from passive observation alone is impossible without reasonable priors, which transformers don’t have. This is why LLMs are brittle and can’t reliably do counter factual.
1
0
3
RT @EliJalbout: 🧵 Let's take a closer look at how the action space impacts #RobotLearning and #SimToReal transfer. 🤖🚀📈🔧 #AI . Blog: https:….
0
8
0
nice example of survivorship bias in the wild: so many guys at faang got a job at faang, so it is not too hard!.
@levelsio I’m an ML scientist at a FAANG company and half my colleagues (including myself) have no CS degree. We transferred from other fields after teaching ourselves ML. Just build cool stuff and then getting hired isn’t too hard.
0
0
1
Should be mandatory for every car.
We deployed 100 deep RL cruise controllers into rush-hour highway traffic to smooth traffic flow and reduce everyone’s energy consumption. Our AVs were decentralized and used standard radar, making our controllers deployable on most cars.1/n
0
0
1