Post-training quantization is widely used to compress large language models (LLMs) for efficient deployment in resource-constrained environments. However, recent work shows that quantization, especially aggressive schemes such as 4-bit QLoRA, can substantially degrade safety alignment, making models more vulnerable to harmful completions and jailbreaks. In this work, we investigate these safety risks and propose a mitigation strategy: projecting quantized parameters back into safety-aligned subspaces. First, we empirically measure safety degradation on benchmark datasets using both safety and utility metrics. Next, we explore projection-based restoration methods to recover alignment-preserving directions in the LoRA adapters of quantized models. Finally, we study how quantization affects mechanistic safety neurons and how hybrid-precision designs can preserve them. By foregrounding the safety implications of model compression, this work aims to support more robust, deployment-ready, and ethically aligned LLMs.

Productivity from day one on supercomputers that leverage new technologies requires significant preparation. An institution that procures a novel system architecture often lacks sufficient institutional knowledge and skills to prepare for it. Thus, the "Center of Excellence" (CoE) concept has emerged to prepare for systems such as Summit and Sierra, currently the top two systems in the Top 500. This paper documents CoE experiences that prepared a workload of diverse applications and math libraries for a heterogeneous system. We describe our approach to this preparation, including our management and execution strategies, and detail our experiences with and reasons for using different programming approaches. Our early science and performance results show that the project enabled significant early seismic science with up to a l4X throughput increase over Cori. In addition to our successes, we discuss our challenges and failures so others may benefit from our experience.

This Innovative Practice Work-In-Progress paper presents a collaborative virtual computer lab (CVCL) environment to support collaborative learning in cloud-based virtual computer labs. With advances of cloud computing and virtualization technologies, a new paradigm of virtual computer labs has emerged, where students carry out labs on virtualized resources remotely through the Internet. Virtual computer labs bring advantages, such as anywhere, anytime, on-demand access of specialized software and hardware. However, with current implementations, it also makes it difficult for students to collaborate, due to the fact that students are assigned separated virtual working spaces in a remote-accessing environment and there is a lack of support for sharing and collaboration. To address this issue, we develop a CVCL environment that allows students to reserve virtual computers labs with multiple participants and support remote real-time collaboration among the participants during a lab. The CVCL environment will implement several well-defined collaborative lab models, including shared remote collaboration, virtual study room, and virtual tutoring center. This paper describes the overall architecture and main features of the CVCL environment and shows preliminary results.