Polynomial Histograms for Memory-Efficient Representation of Long-tailed System Distributions
2026-06-01T08:52:29Z•45949607954e74599e867eaa307750ce4cb3868e4b7a5ef5060fea56a9677169
CRDTCUDA-GraphsDatalogGEMMGPU-performanceInvariant-Bit-PackingLLM-inferencebatch-1distributed-systemsenergy-efficiencyevolutionary-searchfederated-learning','personalization','gumbel-softmax','arxiv','histogramskernel-optimizationlong-tailed-distributionslossless-compressionmemory-bandwidthparallelizationperformance-ruggednessradio-networksschedulingspanning-treetelemetryvirtual-processorwireless-sensor-actuator-networks
What happened
Collection of recent arXiv CS papers (distributed systems, ML systems, and high-performance computing). Highlights: polynomial histograms for compact, low-loss representation of long-tailed telemetry; randomized energy-efficient aggregation and near-optimal minimum-degree spanning-tree construction for radio networks; a Datalog-based declarative framework for specifying and testing CRDT semantics; performance-ruggedness analysis for GEMM with software mitigations that reduce variance and raise throughput; Kernel Foundry, an evolutionary, diagnosis-driven GPU-kernel optimizer with a reusable经验/
Why it matters
A reviewed impact interpretation has not been published for this record.
Evidence and limitations
- Source ID
- arxiv_cs_dc
- Record identifier
- 45949607954e74599e867eaa307750ce4cb3868e4b7a5ef5060fea56a9677169
- Enrichment time
- 2026-06-01T08:52:29Z
- AI-assisted enrichment
- Yes
This record may overlap with other records. Its enrichment can be incomplete or wrong, and machine assistance was used. Validate consequential decisions against the linked source and your own environment.