Gaurav | Searcharxiv

arXiv2026

Twilight of the WIMP: Comprehensive Phenomenology of Electroweak Triplet Dark Matter

We present a comprehensive study of dark matter phenomenology in standard model extensions featuring an electroweak triplet scalar or fermion with hypercharge $Y = 0$ or $Y = 2$. These minimal triplet extensions provide well-motivated dark matter candidates stabilised by the $Z_2$ discrete symmetry. We perform a detailed analysis of the parameter space consistent with current cosmological and experimental constraints, including the relic abundance, direct detection limits, and indirect detection bounds. We find that the scalar triplet with $Y=0$ is ruled out by a combination of relic density, direct detection and indirect detection constraints. On the other hand, the scalar and fermionic triplets with $Y=2$ are both excluded by current direct detection experiments due to their large spin-independent scattering cross-sections. The viable parameter space of the remaining $Y=0$ fermion triplet dark matter lies within the projected sensitivity of near-future experiments, particularly those targeting indirect detection signatures. Collider prospects for these triplet extensions are also discussed in the Appendix.

hep-ph↗

arXiv2021

Efficient Deep Learning Pipelines for Accurate Cost Estimations Over Large Scale Query Workload

The use of deep learning models for forecasting the resource consumption patterns of SQL queries have recently been a popular area of study. With many companies using cloud platforms to power their data lakes for large scale analytic demands, these models form a critical part of the pipeline in managing cloud resource provisioning. While these models have demonstrated promising accuracy, training them over large scale industry workloads are expensive. Space inefficiencies of encoding techniques over large numbers of queries and excessive padding used to enforce shape consistency across diverse query plans implies 1) longer model training time and 2) the need for expensive, scaled up infrastructure to support batched training. In turn, we developed Prestroid, a tree convolution based data science pipeline that accurately predicts resource consumption patterns of query traces, but at a much lower cost. We evaluated our pipeline over 19K Presto OLAP queries from Grab, on a data lake of more than 20PB of data. Experimental results imply that our pipeline outperforms benchmarks on predictive accuracy, contributing to more precise resource prediction for large-scale workloads, yet also reduces per-batch memory footprint by 13.5x and per-epoch training time by 3.45x. We demonstrate direct cost savings of up to 13.2x for large batched model training over Microsoft Azure VMs.

cs.LG↗