4 papers
The Ingestion Tax: Adopting File-Backed Weights in Tensor Frameworks
Yuan Si, Yufeng Lin, Daming Li +1
Open-weight models can occupy a middle capacity regime: active weights fit in DRAM as cached file pages, but a second framework-owned copy does not fit or must be refilled as layer…
Who Should Own the Expert Cache? Kernel-Managed Tiering for Trillion-Parameter MoE Inference
Yuan Si, Yufeng Lin, Daming Li +1
Mixture-of-experts models whose expert pools exceed DRAM capacity require a weight-residency tier. Existing systems manage it in user space with expert-granular placement, frequenc…
Content Hidden Behind Execution: Analyzing Public Scratch Projects at Runtime
Yuan Si, Yufeng Lin, Daming Li +3
Public Scratch projects are reused in computing education as classroom examples, remix sources, open-exploration materials, and research data. Curation often begins with titles, th…
ScratchWorld: Evaluating If World Models Compute Executable Consequences
Yufeng Lin, Jialu Zhang
World-model evaluations often score a predicted future by overlap with a target state or observation. In sparse-change worlds, this can turn copied persistent state into apparent a…