All authors
Claude Skills by hiyenwong
github.com/hiyenwong9,934 skills5 installs19,223 views
- Arxiv 2609 09809v1 Online Inverse Integer Linear Optimization Via Sma**arXiv ID:** 2609.09809v1 **Authors:** Akira Kitaoka **URL:** http://arxiv.org/abs/2609.09809v1 **Utility Score:** 1.00Votes: 0GitHub stars: 3
- Arxiv 2609 09854v1 When Does Low Bit Quantization Preserve The Decisi**arXiv ID:** 2609.09854v1 **Authors:** Wenxuan Xiao, Xu Cao **URL:** http://arxiv.org/abs/2609.09854v1 **Utility Score:** 1.00Votes: 0GitHub stars: 3
- Arxiv 2609 10052v1 Direct Diversity Optimization For Diverse Successf**arXiv ID:** 2609.10052v1 **Authors:** Junwon Ko, Dong-Jae Lee, Minchan Kwon, Sunghyun Baek, Junmo Kim **URL:** http://arxiv.org/abs/2609.10052v1 **Utility Score:** 1.00Votes: 0GitHub stars: 3
- Arxiv 2609 10495v1 Cross Model Agreement As A Deployment Time Reliabi**arXiv ID:** 2609.10495v1 **Authors:** Siddharth Gupta, Jitin Singla **URL:** http://arxiv.org/abs/2609.10495v1 **Utility Score:** 1.00Votes: 0GitHub stars: 3
- Canonical Quantization NeuronsCanonical quantization methodology for constructing quantum neuron models from classical Hamiltonians — a principled framework for quantum machine learning primitivesVotes: 0GitHub stars: 3
- Coflow Scheduling Ocs NetworksScheduling Coflows in Multi-Core Optical Circuit Switching Networks with Performance Guarantee. Optimize parallel data flow coordination in distributed systems using optical circuit switching.Votes: 0GitHub stars: 3
- Coflow Scheduling OcsCoflow Scheduling in Multi-Core Optical Circuit Switching Networks with Performance Guarantee. Optimize parallel data flow coordination in distributed systems using optical circuit switching. Use for data center network optimization, coflow scheduling, and distributed job completion time reduction.Votes: 0GitHub stars: 3
- Comet Constraint Preserving QaoaConstraint-preserving QAOA using XY-mixer for multiplex CRISPR gene editing optimization. Systematically compares structural constraint enforcement (XY-mixer) vs penalty-based approaches across simulator and real hardware. XY-mixer achieves >95% optimum probability by p=3 vs <6% for penalty variants. Activation: constraint-preserving QAOA, XY-mixer, CRISPR optimization, QUBO constraint enforcement, quantum gene editing, 约束保持QAOAVotes: 0GitHub stars: 3
- Deepstruct Linking Deep Learning And Graph Theory**arXiv ID:** 2111.06679 **Authors:** Julian Stier, Michael Granitzer **Published:** 2021-11-12T11:58:13Z **Abstract:** deepstruct connects deep learning models and graph theory such that different graph structures can be imposed on neural networks or graph structures can be extracted from trained neural network models. For this, deepstruct provides deep neural network models with different restrictions which can be created based on an initial graph. Further, tools to extract graph structures...Votes: 0GitHub stars: 3
- Deploymentaligned Lowprecision Neural Architecture Search For Spaceborne Edge Ai**arXiv ID:** 2604.24492 **Authors:** Parampuneet Kaur Thind, Vaibhav Katturu, Giacomo Zema, Roberto Del Prete **Published:** 2026-04-27T13:58:18Z **Abstract:** Designing deep networks that meet strict latency and accuracy constraints on edge accelerators increasingly relies on hardware-aware optimization, including neural architecture search (NAS) guided by device-level metrics. Yet most hardware-aware NAS pipelines still optimize architectures under full-precision assumptions and apply low-...Votes: 0GitHub stars: 3
- Effective Pruning Task Trained Rnn Noisy FluctuationsEffective pruning methodology for task-trained recurrent neural networks using noisy fluctuations and connection rescaling to preserve task performance while maintaining biological plausibility. Implements the noise-prune algorithm that samples connections to preserve based on importance and strengthens retained connections to preserve average synaptic strength. Use when working with recurrent neural networks that need biologically-plausible pruning strategies for functional architectures.Votes: 0GitHub stars: 3
- Effective Pruning Task Trained RnnPrunes RNNs using noise fluctuations and rescaling.Votes: 0GitHub stars: 3
- Eoe Evolutionary Optimization Of Experts For Training Language Models**arXiv ID:** 2509.24436 **Authors:** Yingshi Chen **Published:** 2025-09-29T08:18:26Z **Abstract:** This paper presents an evolutionary framework for the training of large language models(LLM). The models are divided into several experts(sub-networks), which have the same structure but different parameter values. Only one expert is trained at each step. After the classical AdamW optimization, some evolutionary operators(crossover, PSO, and mutation) act on the tensor weights between the curr...Votes: 0GitHub stars: 3
- Evolutionary Architecture Search Through Grammarbased Sequence Alignment**arXiv ID:** 2512.04992 **Authors:** Adri Gómez Martín, Felix Möller, Steven McDonagh, Monica Abella, Manuel Desco, Elliot J. Crowley, Aaron Klein, Linus Ericsson **Published:** 2025-12-04T16:57:49Z **Abstract:** Neural architecture search (NAS) in expressive search spaces is a computationally hard problem, but it also holds the potential to automatically discover completely novel and performant architectures. To achieve this we need effective search algorithms that can identify powerful com...Votes: 0GitHub stars: 3
- Faultaware Design And Training To Enhance Dnns Reliability With Zerooverhead**arXiv ID:** 2205.14420 **Authors:** Niccolò Cavagnero, Fernando Dos Santos, Marco Ciccone, Giuseppe Averta, Tatiana Tommasi, Paolo Rech **Published:** 2022-05-28T13:09:30Z **Abstract:** Deep Neural Networks (DNNs) enable a wide series of technological advancements, ranging from clinical imaging, to predictive industrial maintenance and autonomous driving. However, recent findings indicate that transient hardware faults may corrupt the models prediction dramatically. For instance, the radiat...Votes: 0GitHub stars: 3
- Graphpo Policy Optimization ReasoningGraph-based Policy Optimization (GraphPO) for reasoning models. Represents rollouts as DAGs, merges semantically equivalent paths, and improves advantage estimation variance through graph structure.Votes: 0GitHub stars: 3
- Grinqh Adaptive Quantization HierarchyGraded Input-based Quantization Hierarchy for efficient LLM generation. Dynamic precision assignment based on activation magnitudes as computational importance proxy. Unifies quantization and sparsification.Votes: 0GitHub stars: 3
- Hpc Vqpu ArchitectureService-export architecture for hosting virtual quantum processing units (vQPUs) on batch-scheduled HPC systems. Enables secure supercomputers to expose interactive, backend-oriented quantum interfaces while preserving topology, native-gate, and calibration semantics across queue delays and system scaling.Votes: 0GitHub stars: 3
- Lance Low Rank Activation Compression For Efficient Ondevice Continual Learning**arXiv ID:** 2509.21617 **Authors:** Marco Paul E. Apolinario, Kaushik Roy **Published:** 2025-09-25T21:33:40Z **Abstract:** On-device learning is essential for personalization, privacy, and long-term adaptation in resource-constrained environments. Achieving this requires efficient learning, both fine-tuning existing models and continually acquiring new tasks without catastrophic forgetting. Yet both settings are constrained by high memory cost of storing activations during backpropagation....Votes: 0GitHub stars: 3
- Learning In Feedbackdriven Recurrent Spiking Neural Networks Using Fullforce Training**arXiv ID:** 2205.13585 **Authors:** Ankita Paul, Stefan Wagner, Anup Das **Published:** 2022-05-26T19:01:19Z **Abstract:** Feedback-driven recurrent spiking neural networks (RSNNs) are powerful computational models that can mimic dynamical systems. However, the presence of a feedback loop from the readout to the recurrent layer de-stabilizes the learning mechanism and prevents it from converging. Here, we propose a supervised training procedure for RSNNs, where a second network is introduce...Votes: 0GitHub stars: 3
- Lls Local Learning Rule For Deep Neural Networks Inspired By Neural Activity Synchronization**arXiv ID:** 2405.15868 **Authors:** Marco Paul E. Apolinario, Arani Roy, Kaushik Roy **Published:** 2024-05-24T18:24:24Z **Abstract:** Training deep neural networks (DNNs) using traditional backpropagation (BP) presents challenges in terms of computational complexity and energy consumption, particularly for on-device learning where computational resources are limited. Various alternatives to BP, including random feedback alignment, forward-forward, and local classifiers, have been explored ...Votes: 0GitHub stars: 3
- Minaction Energy First Neural ArchitectureEnergy-first neural architecture design framework based on biological principles. Systematic validation across vision, text, neuromorphic, and physiological datasets with 2,203 experiments. Activation: energy-first, neural architecture, biological principles, energy-regularized, lambda sweep.Votes: 0GitHub stars: 3
- Mocapo Multiobjective Costaware Prompt Optimization**arXiv ID:** 2605.18869 **Authors:** Jan Büssing, Moritz Schlager, Timo Heiß, Tom Zehle, Matthias Feurer **Published:** 2026-05-15T14:56:27Z **Abstract:** Large language models (LLMs) achieve strong performance across a wide range of tasks but are highly sensitive to prompt design, motivating the need for automatic prompt optimization. Existing methods predominantly focus on performance alone, ignoring competing objectives such as inference cost or latency. At the same time, existing work on...Votes: 0GitHub stars: 3
- Neural Circuit Architectural Priors For Quadruped Locomotion**arXiv ID:** 2410.07174 **Authors:** Nikhil X. Bhattasali, Venkatesh Pattabiraman, Lerrel Pinto, Grace W. Lindsay **Published:** 2024-10-09T17:59:45Z **Abstract:** Learning-based approaches to quadruped locomotion commonly adopt generic policy architectures like fully connected MLPs. As such architectures contain few inductive biases, it is common in practice to incorporate priors in the form of rewards, training curricula, imitation data, or trajectory generators. In nature, animals are bor...Votes: 0GitHub stars: 3
- Noisetolerant Coresetbased Class Incremental Continual Learning**arXiv ID:** 2504.16763 **Authors:** Edison Mucllari, Aswin Raghavan, Zachary Alan Daniels **Published:** 2025-04-23T14:34:20Z **Abstract:** Many applications of computer vision require the ability to adapt to novel data distributions after deployment. Adaptation requires algorithms capable of continual learning (CL). Continual learners must be plastic to adapt to novel tasks while minimizing forgetting of previous tasks.However, CL opens up avenues for noise to enter the training pipeline a...Votes: 0GitHub stars: 3
- Ofa2 A Multiobjective Perspective For The Onceforall Neural Architecture Search**arXiv ID:** 2303.13683 **Authors:** Rafael C. Ito, Fernando J. Von Zuben **Published:** 2023-03-23T21:30:29Z **Abstract:** Once-for-All (OFA) is a Neural Architecture Search (NAS) framework designed to address the problem of searching efficient architectures for devices with different resources constraints by decoupling the training and the searching stages. The computationally expensive process of training the OFA neural network is done only once, and then it is possible to perform multipl...Votes: 0GitHub stars: 3
- Orthogonal Residual QuantizationOrthogonal Residual Projection (ORP) framework for multiplier-free Power-of-Two transformer quantization — replaces MAC operations with bit-shifts using dual-basis geometric projection, enabling efficient edge deployment of LLMs at sub-4-bit precision.Votes: 0GitHub stars: 3
- Pendram Enabling Highperformance And Energyefficient Processing Of Deep Neural Networks Through A Generalized Dram Data Mapping Policy**arXiv ID:** 2408.02412 **Authors:** Rachmad Vidya Wicaksana Putra, Muhammad Abdullah Hanif, Muhammad Shafique **Published:** 2024-08-05T12:11:09Z **Abstract:** Convolutional Neural Networks (CNNs), a prominent type of Deep Neural Networks (DNNs), have emerged as a state-of-the-art solution for solving machine learning tasks. To improve the performance and energy efficiency of CNN inference, the employment of specialized hardware accelerators is prevalent. However, CNN accelerators still fac...Votes: 0GitHub stars: 3
- Piper Programmable Distributed TrainingPiper framework for user-controllable distributed training that decouples parallelism strategy from runtime implementation using unified global training DAG intermediate representation. Use for distributed ML training, parallelism strategy design, and flexible training system architecture.Votes: 0GitHub stars: 3
- Posthoc Domain Adaptation Via Guided Data Homogenization**arXiv ID:** 2104.03624 **Authors:** Kurt Willis, Luis Oala **Published:** 2021-04-08T09:18:48Z **Abstract:** Addressing shifts in data distributions is an important prerequisite for the deployment of deep learning models to real-world settings. A general approach to this problem involves the adjustment of models to a new domain through transfer learning. However, in many cases, this is not applicable in a post-hoc manner to deployed models and further parameter adjustments jeopardize safety...Votes: 0GitHub stars: 3
- Pqc Tls DeploymentPost-quantum cryptography TLS deployment methodology. Automated configuration parsing and hybrid PQC (ML-KEM/ML-DSA) deployment patterns for enterprise infrastructure. Use when: deploying post-quantum cryptography, migrating TLS to PQC standards, auditing cryptographic configurations, implementing hybrid key exchanges (MLKEM + X25519), managing crypto-agility in enterprise web server stacks (Nginx, Apache, API gateways), performing cryptographic inventory and exposure assessment, or planning ...Votes: 0GitHub stars: 3
- Predicting And Optimizing For Energy Efficient Acmv Systems Computational Intelligence Approaches**arXiv ID:** 2205.00833 **Authors:** Deqing Zhai, Yeng Chai Soh **Published:** 2022-04-19T09:26:29Z **Abstract:** In this study, a novel application of neural networks that predict thermal comfort states of occupants is proposed with accuracy over 95%, and two optimization algorithms are proposed and evaluated under two real cases (general offices and lecture theatres/conference rooms scenarios) in Singapore. The two optimization algorithms are Bayesian Gaussian process optimization (BGPO) a...Votes: 0GitHub stars: 3
- Pruning Randomly Initialized Neural Networks With Iterative Randomization**arXiv ID:** 2106.09269 **Authors:** Daiki Chijiwa, Shin'ya Yamaguchi, Yasutoshi Ida, Kenji Umakoshi, Tomohiro Inoue **Published:** 2021-06-17T06:32:57Z **Abstract:** Pruning the weights of randomly initialized neural networks plays an important role in the context of lottery ticket hypothesis. Ramanujan et al. (2020) empirically showed that only pruning the weights can achieve remarkable performance instead of optimizing the weight values. However, to achieve the same level of performance a...Votes: 0GitHub stars: 3
- Q Photonas Hybrid Arch SearchQ-PhotoNAS methodology — Hybrid Quantum Neural Architecture Search framework for photonic quantum-classical models using genetic algorithm-based NAS with learnable quantum phase encoding.Votes: 0GitHub stars: 3
- Recent Advances In Federated Learning Driven Large Language Models A Survey On Architecture Performance And Security**arXiv ID:** 2406.09831 **Authors:** Youyang Qu, Ming Liu, Tianqing Zhu, Longxiang Gao, Shui Yu, Wanlei Zhou **Published:** 2024-06-14T08:40:58Z **Abstract:** Federated Learning (FL) offers a promising paradigm for training Large Language Models (LLMs) in a decentralized manner while preserving data privacy and minimizing communication overhead. This survey examines recent advancements in FL-driven LLMs, with a particular emphasis on architectural designs, performance optimization, and secur...Votes: 0GitHub stars: 3
- Recent Advances On Neural Network Pruning At Initialization**arXiv ID:** 2103.06460 **Authors:** Huan Wang, Can Qin, Yue Bai, Yulun Zhang, Yun Fu **Published:** 2021-03-11T05:01:52Z **Abstract:** Neural network pruning typically removes connections or neurons from a pretrained converged model; while a new pruning paradigm, pruning at initialization (PaI), attempts to prune a randomly initialized network. This paper offers the first survey concentrated on this emerging pruning fashion. We first introduce a generic formulation of neural network pruning...Votes: 0GitHub stars: 3
- Refining A Knearest Neighbor Graph For A Computationally Efficient Spectral Clustering**arXiv ID:** 2302.11296 **Authors:** Mashaan Alshammari, John Stavrakakis, Masahiro Takatsuka **Published:** 2023-02-22T11:31:32Z **Abstract:** Spectral clustering became a popular choice for data clustering for its ability of uncovering clusters of different shapes. However, it is not always preferable over other clustering methods due to its computational demands. One of the effective ways to bypass these computational demands is to perform spectral clustering on a subset of points (data r...Votes: 0GitHub stars: 3
- Sageattention2 Efficient Attention With Thorough Outlier Smoothing And Perthread Int4 Quantization**arXiv ID:** 2411.10958 **Authors:** Jintao Zhang, Haofeng Huang, Pengle Zhang, Jia Wei, Jun Zhu, Jianfei Chen **Published:** 2024-11-17T04:35:49Z **Abstract:** Although quantization for linear layers has been widely used, its application to accelerate the attention process remains limited. To further enhance the efficiency of attention computation compared to SageAttention while maintaining precision, we propose SageAttention2, which utilizes significantly faster 4-bit matrix multiplication...Votes: 0GitHub stars: 3
- Selective Task Offloading For Maximum Inference Accuracy And Energy Efficient Realtime Iot Sensing Systems**arXiv ID:** 2402.16904 **Authors:** Abdelkarim Ben Sada, Amar Khelloufi, Abdenacer Naouri, Huansheng Ning, Sahraoui Dhelim **Published:** 2024-02-24T18:46:06Z **Abstract:** The recent advancements in small-size inference models facilitated AI deployment on the edge. However, the limited resource nature of edge devices poses new challenges especially for real-time applications. Deploying multiple inference models (or a single tunable model) varying in size and therefore accuracy and power co...Votes: 0GitHub stars: 3
- Shiftlif Power Of Two QuantizationShiftLIF neuron model: efficient multi-level spiking neurons with power-of-two quantization. Reformulates burst spiking as saturated uniform quantization with learnable scale, absorbable into weights for inference. Activation: shiftlif, power-of-two quantization, learnable quantization scale, multi-level spiking neurons, burst SNN quantization.Votes: 0GitHub stars: 3
- Spike Sparsity Deployment Cost JetsonAnalysis of when spike sparsity in Spiking Neural Networks translates to actual deployment cost savings on commodity edge-GPU platforms. Study of VS-WNO on Jetson Orin Nano with benchmarking methodology.Votes: 0GitHub stars: 3
- Spike Sparsity Deployment CostAnalyzing deployment cost of spiking neural networks on edge hardware. Demonstrates that algorithmic spike sparsity may not translate to actual deployed cost reduction on commodity edge GPUs. Covers VS-WNO (Variable-Spiking Wavelet Neural Operator), WNO comparison, Jetson Orin Nano profiling. Activation: spike sparsity deployment, edge GPU SNN, neuromorphic deployment cost, VS-WNO, wavelet neural operator, SNN hardware, Jetson profiling.Votes: 0GitHub stars: 3
- Spike Sparsity Edge Gpu DeploymentSpike sparsity vs deployed cost analysis for Variable Spiking Wavelet Neural Operator on edge GPU hardware. Activation: brain model, neural scaling, multimodal brain, fMRI, EEG, neural encoding.Votes: 0GitHub stars: 3
- Spikedyn A Framework For Energyefficient Spiking Neural Networks With Continual And Unsupervised Learning Capabilities In Dynamic Environments**arXiv ID:** 2103.00424 **Authors:** Rachmad Vidya Wicaksana Putra, Muhammad Shafique **Published:** 2021-02-28T08:26:23Z **Abstract:** Spiking Neural Networks (SNNs) bear the potential of efficient unsupervised and continual learning capabilities because of their biological plausibility, but their complexity still poses a serious research challenge to enable their energy-efficient design for resource-constrained scenarios (like embedded systems, IoT-Edge, etc.). We propose SpikeDyn, a compr...Votes: 0GitHub stars: 3
- Spiking Manifesto**arXiv ID:** 2512.11843 **Authors:** Eugene Izhikevich **Published:** 2025-12-03T23:44:02Z **Abstract:** Practically everything computers do is better, faster, and more power-efficient than the brain. For example, a calculator performs numerical computations more energy-efficiently than any human. Yet modern AI models are a thousand times less efficient than the brain. These models rely on larger and larger artificial neural networks (ANNs) to boost their encoding capacity, requiring GPUs to...Votes: 0GitHub stars: 3
- Surviving By Serving SbsSurviving by Serving (SBS) principle - functional relevance drives self-organization in complex adaptive systems with multi-agent resource transformationVotes: 0GitHub stars: 3
- The Optimization Trilemma Efficiency Comfort And FDerived from arXiv:2607.17311 - The Optimization Trilemma: Efficiency, Comfort and Fairness in Decentralized Multi-agent CoordinationVotes: 0GitHub stars: 3
- The Robustness Of Spiking Neural Networks In Communication And Its Application Towards Network Efficiency In Federated Learning**arXiv ID:** 2409.12769 **Authors:** Manh V. Nguyen, Liang Zhao, Bobin Deng, William Severa, Honghui Xu, Shaoen Wu **Published:** 2024-09-19T13:37:18Z **Abstract:** Spiking Neural Networks (SNNs) have recently gained significant interest in on-chip learning in embedded devices and emerged as an energy-efficient alternative to conventional Artificial Neural Networks (ANNs). However, to extend SNNs to a Federated Learning (FL) setting involving collaborative model training, the communication b...Votes: 0GitHub stars: 3
- The Shapley Value Of Classifiers In Ensemble Games**arXiv ID:** 2101.02153 **Authors:** Benedek Rozemberczki, Rik Sarkar **Published:** 2021-01-06T17:40:23Z **Abstract:** What is the value of an individual model in an ensemble of binary classifiers? We answer this question by introducing a class of transferable utility cooperative games called \textit{ensemble games}. In machine learning ensembles, pre-trained models cooperate to make classification decisions. To quantify the importance of models in these ensemble games, we define \textit{Tr...Votes: 0GitHub stars: 3
- Vision Smolmamba Token PruningVision SmolMamba: Spike-Guided Token Pruning for energy-efficient spiking state-space vision models. Combines SNN event-driven sparsity with Mamba selective recurrence via SST-TP (Spike-Guided Spatio-Temporal Token Pruner). Activation: Vision SmolMamba, spike-guided token pruning, state-space SNN, Mamba vision, SST-TPVotes: 0GitHub stars: 3