Omead Pooladzandi ✈️ NeurIPS'24
hessianfree.bsky.social
Optimization Generative Modeling @Caltech, PhD @UCLA. ex Research Scientist Intern @AIatMeta (opinions are my own) why is jax so difficult
👤 Joined on 21 November 2024
477
Followers
820
Following
26
Total Posts
📜 Kullanıcı Geçmişi
- 21 Kas 2024, 04:54
🎉 Bluesky'e katıldı - İlk kullanıcı adı:
@hessianfree.bsky.social
🚫 Blokladığı Kullanıcılar
❤️ Beğendiği Gönderiler
Pretraining will only end once we find the optimal scaling law.
Show Post
It being physically small and very low power consumption at idle probably also plays a role (vs having a desktop computer with 2 4090s idling)
Show Postthe o3 announcement feels very analogous to the GPT-3 announcement in early 2020: "here is new tech, it's very powerful, but it's super expensive to run and we're super selective with who we give access to"
Show Postwho remembers GPT-3 Davinci pricing?
Show Post
Probably my biggest naming blunder was the name "auto-encoding variational Bayes" for a method that (in its default version) is only Bayesian over the latent variables, not the parameters. Blasphemy in the church it Bayes 😂
Show Post