Conference Paper
Recent Submissions
-
The Hydrogen Intensity Real-time Analysis eXperiment: Overview and Status Update
(2024)2024 18TH EUROPEAN CONFERENCE ON ANTENNAS AND PROPAGATION, EUCAPThe Hydrogen Intensity Real-time Analysis eXperiment (HIRAX) will be a large interferometric array of drift-scan radio telescopes designed to map the large-scale spatial fluctuations of neutral hydrogen in the Universe, in order to better understand the nature of dark energy. It will operate between 400-800 MHz, and is currently under construction in the Karoo desert of South Africa. It will also be a powerful tool for studying astronomical ...Conference Paper -
Validating the eBPF Verifier via State Embedding
(2024)PROCEEDINGS OF THE 18TH USENIX SYMPOSIUM ON OPERATING SYSTEMS DESIGN AND IMPLEMENTATION, OSDI 2024This paper introduces state embedding, a novel and highly effective technique for validating the correctness of the eBPF verifier, a critical component for Linux kernel security. To check whether a program is safe to execute, the verifier must track over-approximated program states along each potential control-flow path; any concrete state not contained in the tracked approximation may invalidate the verifier's conclusion. Our key insight ...Conference Paper -
ProteinShake Building datasets and benchmarks for deep learning on protein structures
(2023)Advances in Neural Information Processing SystemsWe present ProteinShake, a Python software package that simplifies dataset creation and model evaluation for deep learning on protein structures. Users can create custom datasets or load an extensive set of pre-processed datasets from biological data repositories such as the Protein Data Bank (PDB) and AlphaFoldDB. Each dataset is associated with prediction tasks and evaluation functions covering a broad array of biological challenges. A ...Conference Paper -
Detecting Logic Bugs in Database Engines via Equivalent Expression Transformation
(2024)PROCEEDINGS OF THE 18TH USENIX SYMPOSIUM ON OPERATING SYSTEMS DESIGN AND IMPLEMENTATION, OSDI 2024Database management systems (DBMSs) are crucial for storing and fetching data. To improve the reliability of such systems, approaches have been proposed to detect logic bugs that cause DBMSs to process data incorrectly. These approaches manipulate queries and check whether the query results produced by DBMSs follow the expectations. However, such query-level manipulation cannot handle complex query semantics and thus needs to limit the ...Conference Paper -
Pecan: Cost-Efficient ML Data Preprocessing with Automatic Transformation Ordering and Hybrid Placement
(2024)PROCEEDINGS OF THE 2024 USENIX ANNUAL TECHNICAL CONFERENCE, ATC 2024Input data preprocessing is a common bottleneck in machine teaming (ML) jobs, that can significantly increase training time and cost as expensive GPUs or Till's idle waiting for input data. Previous Work has shown that offloading data preprocessing to remote CPU servers successfully alleviates data stalls and improves training time. However, remote CPU workers in disaggregated data processing systems comprise a significant fraction of ...Conference Paper -
OSMOSIS: Enabling Multi-Tenancy in Datacenter SmartNICs
(2024)PROCEEDINGS OF THE 2024 USENIX ANNUAL TECHNICAL CONFERENCE, ATC 2024Multi-tenancy is essential for unleashing SmartNIC's potential in datacenters. Our systematic analysis in this work shows that existing on-path SmartNICs have resource multiplexing limitations. For example, existing solutions lack multi-tenancy capabilities such as performance isolation and QoS provisioning for compute and 10 resources. Compared to standard NIC data paths with a well-defined set of offloaded functions, unpredictable ...Conference Paper -
Hierarchical Integration Diffusion Model for Realistic Image Deblurring
(2023)Advances in Neural Information Processing SystemsDiffusion models (DMs) have recently been introduced in image deblurring and exhibited promising performance, particularly in terms of details reconstruction. However, the diffusion model requires a large number of inference iterations to recover the clean image from pure Gaussian noise, which consumes massive computational resources. Moreover, the distribution synthesized by the diffusion model is often misaligned with the target results, ...Conference Paper -
Goal-conditioned Offline Planning from Curious Exploration
(2023)Advances in Neural Information Processing SystemsCuriosity has established itself as a powerful exploration strategy in deep reinforcement learning. Notably, leveraging expected future novelty as intrinsic motivation has been shown to efficiently generate exploratory trajectories, as well as a robust dynamics model. We consider the challenge of extracting goal-conditioned behavior from the products of such unsupervised exploration techniques, without any additional environment interaction. ...Conference Paper -
GEO-Bench: Toward Foundation Models for Earth Monitoring
(2023)Advances in Neural Information Processing SystemsRecent progress in self-supervision has shown that pre-training large neural networks on vast amounts of unsupervised data can lead to substantial increases in generalization to downstream tasks. Such models, recently coined foundation models, have been transformational to the field of natural language processing. Variants have also been proposed for image data, but their applicability to remote sensing tasks is limited. To stimulate the ...Conference Paper -
Human-Aligned Calibration for AI-Assisted Decision Making
(2023)Advances in Neural Information Processing SystemsWhenever a binary classifier is used to provide decision support, it typically provides both a label prediction and a confidence value. Then, the decision maker is supposed to use the confidence value to calibrate how much to trust the prediction. In this context, it has been often argued that the confidence value should correspond to a well calibrated estimate of the probability that the predicted label matches the ground truth label. ...Conference Paper -
Efficiently Computing Directed Minimum Spanning Trees
(2023)2023 PROCEEDINGS OF THE SYMPOSIUM ON ALGORITHM ENGINEERING AND EXPERIMENTS, ALENEXComputing a directed minimum spanning tree, called arborescence, is a fundamental algorithmic problem, although not as common as its undirected counterpart. In 1967, Edmonds discussed an elegant solution. It was refined to run in O(min(n(2), mlog n)) by Tarjan which is optimal for very dense and very sparse graphs. Gabow et al. gave a version of Edmonds' algorithm that runs in O(n log n+ m), thus asymptotically beating the Tarjan variant ...Conference Paper -
Low-Depth Spatial Tree Algorithms
(2024)International Parallel and Distributed Processing Symposium IPDPSContemporary accelerator designs exhibit a high degree of spatial localization, wherein two-dimensional physical distance determines communication costs between processing elements. This situation presents considerable algorithmic challenges, particularly when managing sparse data, a pivotal component in progressing data science. The spatial computer model quantifies communication locality by weighting processor communication costs by ...Conference Paper -
Software Resource Disaggregation for HPC with Serverless Computing
(2024)International Parallel and Distributed Processing Symposium IPDPSAggregated HPC resources have rigid allocation systems and programming models which struggle to adapt to diverse and changing workloads. Consequently, HPC systems fail to efficiently use the large pools of unused memory and increase the utilization of idle computing resources. Prior work attempted to increase the throughput and efficiency of super-computing systems through workload co-location and resource disaggregation. However, these ...Conference Paper -
Towards Efficient Confluent Edge Networks
(2024)2024 JOINT EUROPEAN CONFERENCE ON NETWORKS AND COMMUNICATIONS & 6G SUMMIT, EUCNC/6G SUMMIT 2024We provide a vision for 6G fixed networks based on flexible and scalable high-capacity transmission technologies that form mesh edge networks to achieve ultra energy-efficient highly available networks with low latency. These networks will be controlled by AI-native orchestration across mobile, fixed, and compute domains. Mesh networking at the edge will be enabled by a seamless 'confluence' of radio fixed wireless (RFW), free space optical ...Conference Paper -
Challenges and Opportunities in Text Generation Explainability
(2024)EXPLAINABLE ARTIFICIAL INTELLIGENCE, PT I, XAI 2024The necessity for interpretability in natural language processing (NLP) has risen alongside the growing prominence of large language models. Among the myriad tasks within NLP, text generation stands out as a primary objective of autoregressive models. The NLP community has begun to take a keen interest in gaining a deeper understanding of text generation, leading to the development of model-agnostic explainable artificial intelligence ...Conference Paper -
A visual approach to structural design: photoelasticity as a collaborative tool in Gengo Matsui’s work
(2024)Construction Matters: Proceedings of the 8th International Congress on Construction HistoryThis paper focuses on the role of photoelasticity as a design tool in the work of the Japanese structural engineer Gengo Matsui (1920–1996). The objective is to shed light on the operative dimension of these experiments, investigating their role as part of a holistic design process to which both the architect and the structural engineer actively contribute. The paper reconstructs the role of photoelasticity through a comprehensive literature ...Conference Paper -
Basement V4—A Multipurpose Modelling Environment for Simulation of Flood Hazards and River Morphodynamics Across Scales
(2024)Advances in Hydroinformatics - SimHydro 2023: New Modelling Paradigms for Water IssuesThe numerical modelling of hydro- and morphodynamics in watercourses is of great importance for both the advancement of scientific understanding of under lying processes and for design in engineering practice. For this purpose, and focusing on alpine and subalpine conditions, the modelling software BASEMENT has been developed in the last two decades. The simulation environment is currently composed by different tools, either freeware or ...Conference Paper -
Scaling and performance portability of the particle-in-cell scheme for plasma physics applications through mini-apps targeting exascale architectures
(2024)PROCEEDINGS OF THE 2024 SIAM CONFERENCE ON PARALLEL PROCESSING FOR SCIENTIFIC COMPUTING, PPWe perform a scaling and performance portability study of the electrostatic particle-in-cell scheme for plasma physics applications through a set of mini-apps we name "Alpine", which can make use of exascale computing capabilities. The mini-apps are based on IPPL, a framework that is designed around performance portable and dimensionality independent particles and fields. We benchmark the simulations with varying parameters, such as grid ...Conference Paper -
Re-interpreting the Step-Response Probability Curve to Extract Fundamental Physical Parameters of Event-Based Vision Sensors
(2024)UNCONVENTIONAL OPTICAL IMAGING IVBiologically inspired event-based vision sensors (EVS) are growing in popularity due to performance benefits including ultra-low power consumption, high dynamic range, data sparsity, and fast temporal response. They efficiently encode dynamic information from a visual scene through pixels that respond autonomously and asynchronously when the per-pixel illumination level changes by a user-selectable contrast threshold ratio,. theta Due to ...Conference Paper -
On Structure-Preserving Cryptography and Lattices
(2024)PUBLIC-KEY CRYPTOGRAPHY, PT III, PKC 2024The Groth-Sahai proof system is a highly efficient pairing-based proof system for a specific class of group-based languages. Cryptographic primitives that are compatible with these languages (such that we can express, e.g., that a ciphertext contains a valid signature for a given message) are called "structure-preserving". The combination of structure-preserving primitives with Groth-Sahai proofs allows to prove complex statements that ...Conference Paper