The dynamic allocation of spectrum in 5G / 6G networks is critical to efficient resource utilization. However, applying traditional deep reinforcement learning (DRL) is often infeasible due to its immense sample complexity and the safety risks associ...
We propose a class of strongly efficient rare event simulation estimators for random walks and compound Poisson processes with a regularly varying increment/jump-size distribution in a general large deviations regime. Our estimator is based on an imp...
Code library that uses S3's API to provide an efficient random-access (seekable) Stream implementation for use in code where efficient network I/O is paramount. (⭐ 43)
In this work, we present AfriNLLB, a series of lightweight models for efficient translation from and into African languages. AfriNLLB supports 15 language pairs (30 translation directions), including Swahili, Hausa, Yoruba, Amharic, Somali, Zulu, Lin...
Parameter-Efficient Fine-Tuning (PEFT) effectively adapts pre-trained transformers to downstream tasks. However, the optimization of tasks performance often comes at the cost of generalizability in fine-tuned models. To address this issue, we theoret...
In this paper, we propose a binarized neural network learning method called BiDet for efficient object detection. Conventional network binarization methods directly quantize the weights and activations in one-stage or two-stage detectors with constra...
The plasma wave instabilities in ballistic Field Effect Transistors (FETs) have a promise of developing sensitive THz detectors and efficient THz sources. One of the difficulties in achieving efficient resonant plasmonic detection and generation is a...
by efficient voice leadings. Clearly, to satisfy these constraints, a composer must be able to find the efficient voice leadings between arbitrary chords
We address the task of video chaptering, i.e., partitioning a long video timeline into semantic units and generating corresponding chapter titles. While relatively underexplored, automatic chaptering has the potential to enable efficient navigation a...
Efficiently implemented Kennard-Stone sampling algorithms for quite large datasets (⭐ 5)
We present a simple, efficient model for learning boundary detection based on a random forest classifier. Our approach combines (1) efficient clustering of training examples based on simple partitioning of the space of local edge orientations and (2)...
While the tree-based machine learning (TBML) models exhibit superior performance compared to neural networks on tabular data and hold promise for energy-efficient acceleration using aCAM arrays, their ideal deployment on hardware with explicit exploi...
Efficient inference of LLMs remains a crucial challenge, with two main phases: a compute-intensive prompt computation and a memory-intensive token generation. Despite existing batching and scheduling techniques, token generation phases fail to fully...
What is the most efficient way for the human body to transport itself without using any artificial enhancements?...
We present LazyDINO, a transport map variational inference method for fast, scalable, and efficiently amortized solutions of high-dimensional nonlinear Bayesian inverse problems with expensive parameter-to-observable (PtO) maps. Our method consists o...
Dense linear layers are the dominant computational bottleneck in foundation models. Identifying more efficient alternatives to dense matrices has enormous potential for building more compute-efficient models, as exemplified by the success of convolut...
This study aims to explore efficient tuning methods for the screenshot captioning task. Recently, image captioning has seen significant advancements, but research in captioning tasks for mobile screens remains relatively scarce. Current datasets and...
Geometry optimization is efficient using generalized Gaunt coefficients, which significantly limit the amount of cross differentiation for multi-center integrals of high-angular-momentum solid-harmonic basis sets. We parameterize the fully analytic...
This note describes a new efficient bit-slice implementation DenseQMC of the Quine-McCluskey algorithm for finding all prime implicants of a Boolean function in the dense case. It is practically feasible for n <= 23 when run on a common laptop or for...
In this paper we propose efficient methods for elicitation of complexly structured preferences and utilize these in problems of decision making under (severe) uncertainty. Based on the general framework introduced in Jansen, Schollmeyer and Augustin...
The massive adoption of large language models (LLMs) demands efficient deployment strategies. However, the auto-regressive decoding process, which is fundamental to how most LLMs generate text, poses challenges to achieve efficient serving. In this w...
A new set of hardware merge sort devices are introduced here, which merge multiple sorted input lists into a single sorted output list in a fast and efficient manner. In each merge sorter, the values from the sorted input lists are arranged in an inp...
In this contribution we analyze the effect that mutual information has on the actual performance of efficient negotiation methods. Specifically, we start by proposing the theoretical notion of Abstract Negotiation Method (ANM) as a map from the negot...
No description (⭐ 1008)
Tags: c, algorithm, assembly, bit-manipulation, micro-optimization | Score: 20
Tags: c++, file-io | Score: 356
This is a technical report, containing all the theorem proofs and additional evaluations in paper "Efficient Identification of Additive Link Metrics via Network Tomography" by Liang Ma, Ting He, Kin K. Leung, Don Towsley, and Ananthram Swami, publish...
Tags: c#, regex, performance | Score: 1352
Tags: python, numpy | Score: 380
We present an efficient method of pretraining large-scale autoencoding language models using training signals generated by an auxiliary model. Originated in ELECTRA, this training strategy has demonstrated sample-efficiency to pretrain models at the...
The mining of bitcoin is modeled using system dynamics, showing that the past evolution of the network hash rate can be explained to a large extent by an efficient market hypothesis applied to the mining of blocks. The possibility of a decrease in th...
In this paper, we investigate the capability of the universal Kriging (UK) model for single-objective global optimization applied within an efficient global optimization (EGO) framework. We implemented this combined UK-EGO framework and studied four...
Recently, Younes et al. proposed an efficient multi-party semi-quantum secret sharing (SQSS) scheme that generalizes Tian et al.'s three-party protocol \cite{Tian2021} to accommodate multiple participants. This scheme retains the original advantages,...
The concept of efficiency plays a prominent role in the formal solution of decision problems that involve incomparable alternatives. This paper develops necessary and sufficient conditions for the efficient points in a sum of sets of alternatives to...
Motivated by the fact that entities in a social network or biological system often interact by exchanging information, we propose an efficient info-clustering algorithm that can group entities into communities using a parametric max-flow algorithm. T...
In the age of information overload, content management for online news articles relies on efficient summarization to enhance accessibility and user engagement. This article addresses the challenge of extractive text summarization by employing advance...
Processing very large graphs like social networks, biological and chemical compounds is a challenging task. Distributed graph processing systems process the billion-scale graphs efficiently but incur overheads of efficient partitioning and distributi...
Nov 22, 2024 · MIT researchers developed an efficient approach for training more reliable reinforcement learning models, focusing on complex tasks that involve variability. This could enable the leverage of …
Long-context language models exhibit impressive performance but remain challenging to deploy due to high GPU memory demands during inference. We propose Memory-efficient Offloaded Mini-sequence Inference (MOM), a method that partitions critical layer...
Many immunization strategies have been proposed to prevent infectious viruses from spreading through a network. In this study, we propose efficient immunization strategies to prevent a default contagion that might occur in a financial network. An ess...