BRAInLab

Basic Research of AIn Lab: optimization in ML, federated & distributed learning, convex & stochastic optimization theory, privacy & trusted learning

[BRAInLab @Lab4BRAIn]

Research this year at ICLR with 5 accepted papers 🎉

If you’re in Brazil — come meet the team, we’d love to connect!

[BRAInLab @Lab4BRAIn]

BRAIn Lab × MBZUAI — 5 years of collaboration 🔥

Together, we’ve prepared a survey on Byzantine attacks — link below 👇

https://t.co/sNGzIN0K5L

Despite a challenging period in Abu Dhabi, the work continues. Supporting our colleagues 🙏

[BRAInLab @Lab4BRAIn]

AAAI-26 — done ✅

Emotions overflowing, great scientific connections, and tons of ideas. Next up — ICLR 👀✨

[BRAInLab @Lab4BRAIn]

To get more details👇🏻

📄 Paper: https://t.co/mOhT8fwWEi

🖇️ References:
[1] https://t.co/rxHD8xjE0e
[2] https://t.co/WoVDUw4qLi
[3] https://t.co/Anw3bZp5ox

[BRAInLab @Lab4BRAIn]

Adaptive step sizes without knowing the smoothness constant? That’s real! Compatible with ANY variance-reduction method — finite-sum, distributed, coordinate sketches included

📜 Unified Theory of Adaptive Variance Reduction

From pure SGD to advances like variance reduction

[BRAInLab @Lab4BRAIn]

As a result, these adaptive step sizes behave exactly like the tuned ones — or slightly better

[BRAInLab @Lab4BRAIn]

Ready to dive into the details or try it yourself? 🚀

📄 Paper: https://t.co/h98ALg3qU6
💻 Code: https://t.co/2Zg5FZpRYw

[BRAInLab @Lab4BRAIn]

Struggling with data heterogeneity in your deep learning projects? We introduce ALSO, a new, practical Distributionally Robust Optimizer (DRO)🚀

📜 Paper thread for "Aligning Distributionally Robust Optimization with Practical Deep Learning Needs"

[BRAInLab @Lab4BRAIn]

We tested ALSO across diverse setups with real-world data heterogeneity, including:
• Class-imbalanced datasets
• Tabular DL
• Robust & distributed training
• Split learning
In all scenarios, ALSO showed superior performance over both standard approaches and DRO baselines