Claas Voelcker
@cvoelcker.bsky.social
2.6K followers 390 following 540 posts
Thou shalt not disfigure the soul! The square root of a goat is one If I seem very angry, check if I have been watered in the last 24 hours. For professional, see https://cvoelcker.de
Posts Media Videos Starter Packs
Pinned
cvoelcker.bsky.social
And most importantly, I'm at least 200% nicer than I appear in any given moment, so let me know if you really just want to geek out about random shit!
cvoelcker.bsky.social
I'm Claas, a hopefully-soon-finished-PhD researcher at @uoft.bsky.social . I work on reinforcement learning, especially on deep model-based methods, and am dabbling in diffusion policies and large-scale imitation learning.
I'm way too political and loud in general, so please be warned.
cvoelcker.bsky.social
what did twitter do now?
cvoelcker.bsky.social
I mean, don't we have high school tracks now :D
cvoelcker.bsky.social
i literally had an algorithm idea come to me in a fever dream, but it works???
cvoelcker.bsky.social
value functions are a mirage
Reposted by Claas Voelcker
sologen.bsky.social
Apply to join one of the top AI research 🤖 institutes in the world in the wonderful city of Montréal!
I will be recruiting 1-2 students this year. Just mention my name in your application.
mila-quebec.bsky.social
Mila's annual supervision request process is now open to receive MSc and PhD applications for Fall 2026 admission! For more information, visit mila.quebec/en/prospecti...
cvoelcker.bsky.social
Honestly, theory people would be so much more celebrated and cited if they didn't insist on removing _all_ intuition from their papers about why their methods work :D Sooooooooo much labour in reverse engineering the intuitions, which often turn out to be pretty simple.
cvoelcker.bsky.social
Every class can be a Barbarian if you just play it like one!
cvoelcker.bsky.social
Since there is a decent amount of "Do we know how LLMs work?" discourse flowing around, I would be really interested to hear what people would accept as "understanding how LLMs work". What kind of knowledge are we speaking about here?
cvoelcker.bsky.social
Sometimes D&D adventures are hilariously badly written:

Players hear a voice telling them to bring NPC X. They have to bring NPC X to open the door. If NPC X died earlier in the adventure, X will wait for them behind the door.

So if X dies it is behind the door that can ONLY be opened by X???
Reposted by Claas Voelcker
eleanor.lockhart.contact
A lot of folks in the AI space are like people who insist it’s safe and smart to get drunk in the driver’s seat of their Tesla on a country road and a lot of their opponents are like people saying there’s no such thing as an automatic transmission
cvoelcker.bsky.social
I don't understand how the Canadian AI/ML ecosystem wants to attract and retain talents when they are offering less then half the salary on representative roles...
cvoelcker.bsky.social
Put on arxiv before acceptance, yes or no?
cvoelcker.bsky.social
Really great rant!
davidpicard.bsky.social
It's a great discussion because we all have different interpretation of what this means and what the consequences are.

Myself, I have absolutely no doubt that scaling works. If you have all the videos in the world and are able to train a model that can recall and merge any of them, then for sure...
Reposted by Claas Voelcker
spacecowboy17.bsky.social
Enjoying the For You feed? Give it a like ♡ to help more people discover it: bsky.app/profile/did:...

The more people use it -> the more feedback we get -> the better we can make it for you.
cvoelcker.bsky.social
A totally unrelated question: does anybody know how to make long equations work on mobile with math jax and Jekyll 😅🙈
cvoelcker.bsky.social
Congrats! May your GPU and space access live long and prosper!
Reposted by Claas Voelcker
eugenevinitsky.bsky.social
RL rant time after reading another LLM paper about whether "RL sharpens the distribution or discovers new knowledge.": RL is not magic. If your exploration policy takes an action with 0 probability, it can't explore that action! It trivially just affects the distribution of supported actions.
cvoelcker.bsky.social
Stanford and Berkeley are functionally equivalent places and I refuse to treat them as separate entities.
cvoelcker.bsky.social
But those are WOOOOOOORK :D
cvoelcker.bsky.social
My grad school salary advise: find a loving partner before grad school, get them a work visa and a well-paid job in your school location, have them wildly out-earn you because they are brilliant and tada... sugar-partner! Tested for your convenience!
eugenevinitsky.bsky.social
My career advice? Come into a big inheritance from a long lost aunt as well as a big, possibly haunted house formerly inhabited by generations of your family
cvoelcker.bsky.social
I pitched this ~8 years ago! It was going to provide the services "AI consultancy" and "blockchain consultancy". Our highly trained consultants would say "no" when you ask about AI or blockchain, and then just give you a normal database and some working SQL.
markriedl.bsky.social
I’m starting a new AI consulting company called It’s Not Going To Work, LLC. Here’s how it works. You give me $100,000, I tell you “It’s not going to work”(TM), you save $1M+ in wasted contracts, lots of wasted effort, and embarrassment. It’s a good deal. Try it out.