Dan Alistarh's Website (IST Austria Distributed Algorithms and Systems Lab) (daslab.ista.ac.at)

Dan Alistarh, a Professor at the Institute of Science and Technology Austria (ISTA) and ML Research Lead at Neural Magic, has announced a significant number of research papers accepted at prominent AI conferences, further solidifying the lab's influence in the machine learning community. The group secured five accepted papers for NeurIPS 2025, including "HALO," which focuses on the accurate quantization of INT8 training for general models, and "Compression Scaling Laws," essential for understanding scaling behavior in compressed ML models. Additionally, their work has been recognized at ICML, PPoPP, and EMNLP, showcasing the lab's commitment to advancing efficient algorithms and systems in ML. This prolific output underscores the lab's ongoing dedication to improving the efficiency of machine learning techniques, particularly in quantization and sparsity, which are crucial for enhancing model performance while reducing resource consumption. The implications of this research extend to various applications, especially in deploying large language models (LLMs) in resource-constrained environments. With strong industry support and a focus on optimization theory and distributed systems, Alistarh's lab is poised to make further impactful contributions to the field, especially as they seek new talent through open positions.