Tootfinder

I really liked Marvel's Midnight Suns, and it's free on the EGS right now, for a few days: https://store.epicgames.com/en-US/p/marvels-midnight-suns

This https://arxiv.org/abs/2405.08781 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_mat…

Total coloring and edge-girth coloring of regular graphs via efficient dominating sets or perfect codes
A total coloring of a regular graph $G$ via a color set formed by one more unit than the degree of $G$ is said to be a totally efficient coloring if each color class is an efficient dominating set (or perfect code). We show that the binary 3-cube graph has a totally efficient coloring that in turn yields an edge-girth coloring of the binary 4-cube. We conjecture that this graph $G$ is the only existing case of a totally efficient coloring and subsequent edge-girth coloring of the prism of $G$. …

Efficient Leverage Score Sampling for Tensor Train Decomposition
Vivek Bharadwaj, Beheshteh T. Rakhshan, Osman Asif Malik, Guillaume Rabusseau
https://arxiv.org/abs/2406.02749

Efficient Leverage Score Sampling for Tensor Train Decomposition
Tensor Train~(TT) decomposition is widely used in the machine learning and quantum physics communities as a popular tool to efficiently compress high-dimensional tensor data. In this paper, we propose an efficient algorithm to accelerate computing the TT decomposition with the Alternating Least Squares (ALS) algorithm relying on exact leverage scores sampling. For this purpose, we propose a data structure that allows us to efficiently sample from the tensor with time complexity logarithmic in t…

Seq1F1B: Efficient Sequence-Level Pipeline Parallelism for Large Language Model Training
Sun Ao, Weilin Zhao, Xu Han, Cheng Yang, Zhiyuan Liu, Chuan Shi, Maosong Sun
https://arxiv.org/abs/2406.03488

Seq1F1B: Efficient Sequence-Level Pipeline Parallelism for Large Language Model Training
The emergence of large language models (LLMs) relies heavily on distributed training strategies, among which pipeline parallelism plays a crucial role. As LLMs' training sequence length extends to 32k or even 128k, the current pipeline parallel methods face severe bottlenecks, including high memory footprints and substantial pipeline bubbles, greatly hindering model scalability and training throughput. To enhance memory efficiency and training throughput, in this work, we introduce an efficient…

This https://arxiv.org/abs/2402.07844 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csSE_…

Mercury: A Code Efficiency Benchmark for LLM Code Synthesis
Amidst the recent strides in evaluating Large Language Models for Code (Code LLMs), existing benchmarks have mainly focused on the functional correctness of generated code, neglecting the importance of their computational efficiency. To fill the gap, we present Mercury, the first code efficiency benchmark for Code LLMs. It comprises 1,889 Python tasks, each accompanied by adequate solutions that serve as real-world efficiency baselines, enabling a comprehensive analysis of the runtime distribut…

Balancing Performance and Efficiency in Zero-shot Robotic Navigation
Dmytro Kuzmenko, Nadiya Shvai
https://arxiv.org/abs/2406.03015 https://

Balancing Performance and Efficiency in Zero-shot Robotic Navigation
We present an optimization study of the Vision-Language Frontier Maps (VLFM) applied to the Object Goal Navigation task in robotics. Our work evaluates the efficiency and performance of various vision-language models, object detectors, segmentation models, and multi-modal comprehension and Visual Question Answering modules. Using the $\textit{val-mini}$ and $\textit{val}$ splits of Habitat-Matterport 3D dataset, we conduct experiments on a desktop with limited VRAM. We propose a solution that a…

Energy-Efficient Hybrid Beamforming for Integrated Sensing and Communication Enabled mmWave MIMO Systems
Jitendra Singh, Suraj Srivastava, Aditya K. Jagannatham
https://arxiv.org/abs/2406.03737

Energy-Efficient Hybrid Beamforming for Integrated Sensing and Communication Enabled mmWave MIMO Systems
This paper conceives a hybrid beamforming design (HBF) that maximizes the energy efficiency (EE) of an integrated sensing and communication (ISAC)-enabled millimeter wave (mmWave) multiple-input multiple-output (MIMO) system. In the system under consideration, an ISAC base station (BS) with the hybrid MIMO architecture communicates with multiple users and simultaneously detects multiple targets. The proposed scheme seeks to maximize the EE of the system, considering the signal-to-interference a…

Balancing Performance and Efficiency in Zero-shot Robotic Navigation
Dmytro Kuzmenko, Nadiya Shvai
https://arxiv.org/abs/2406.03015 https://

Balancing Performance and Efficiency in Zero-shot Robotic Navigation
We present an optimization study of the Vision-Language Frontier Maps (VLFM) applied to the Object Goal Navigation task in robotics. Our work evaluates the efficiency and performance of various vision-language models, object detectors, segmentation models, and multi-modal comprehension and Visual Question Answering modules. Using the $\textit{val-mini}$ and $\textit{val}$ splits of Habitat-Matterport 3D dataset, we conduct experiments on a desktop with limited VRAM. We propose a solution that a…

This https://arxiv.org/abs/2406.02749 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csDS_…

Efficient Leverage Score Sampling for Tensor Train Decomposition
Tensor Train~(TT) decomposition is widely used in the machine learning and quantum physics communities as a popular tool to efficiently compress high-dimensional tensor data. In this paper, we propose an efficient algorithm to accelerate computing the TT decomposition with the Alternating Least Squares (ALS) algorithm relying on exact leverage scores sampling. For this purpose, we propose a data structure that allows us to efficiently sample from the tensor with time complexity logarithmic in t…

This https://arxiv.org/abs/2406.03488 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csDC_…

Seq1F1B: Efficient Sequence-Level Pipeline Parallelism for Large Language Model Training
The emergence of large language models (LLMs) relies heavily on distributed training strategies, among which pipeline parallelism plays a crucial role. As LLMs' training sequence length extends to 32k or even 128k, the current pipeline parallel methods face severe bottlenecks, including high memory footprints and substantial pipeline bubbles, greatly hindering model scalability and training throughput. To enhance memory efficiency and training throughput, in this work, we introduce an efficient…