All authors
Claude Skills by hiyenwong
github.com/hiyenwong9,934 skills5 installs19,223 views
- Openai Research MonitorSkill for AI agent capabilitiesVotes: 0GitHub stars: 3
- Optimizeanything A Universal Api For Optimizing Any Text Parameter**arXiv ID:** 2605.19633 **Authors:** Lakshya A Agrawal, Donghyun Lee, Shangyin Tan, Wenjie Ma, Karim Elmaaroufi, Rohit Sandadi, Sanjit A. Seshia, Koushik Sen, Dan Klein, Ion Stoica, Joseph E. Gonzalez, Omar Khattab, Alexandros G. Dimakis, Matei Zaharia **Published:** 2026-05-19T10:18:12Z **Abstract:** Can a single LLM-based optimization system match specialized tools across fundamentally different domains? We show that when optimization problems are formulated as improving a text artifact ev...Votes: 0GitHub stars: 3
- Pals Percentile Aware Layerwise Sparsity For Llm PruningOne-shot pruning methods like Wanda and SparseGPT apply the same sparsity ratio to every layer of a transformer, ignoring known variation in layer importance. We propose PALS (Percentile-Aware Layerwi. Based on arXiv:2607.07557.Votes: 0GitHub stars: 3
- Paper Llm EvalSkill for AI agent capabilitiesVotes: 0GitHub stars: 3
- Parameterefficient Finetuning Of Llms With Mixture Of Space Experts**arXiv ID:** 2602.14490 **Authors:** Buze Zhang, Jinkai Tao, Zilang Zeng, Neil He, Ali Maatouk, Menglin Yang, Rex Ying **Published:** 2026-02-16T06:07:32Z **Abstract:** Large Language Models (LLMs) have achieved remarkable progress, with Parameter-Efficient Fine-Tuning (PEFT) emerging as a key technique for downstream task adaptation. However, existing PEFT methods mainly operate in Euclidean space, fundamentally limiting their capacity to capture complex geometric structures inherent in lan...Votes: 0GitHub stars: 3
- Perceptioninformed Neural Networks Beyond Physicsinformed Neural Networks**arXiv ID:** 2505.03806 **Authors:** Mehran Mazandarani, Marzieh Najariyan **Published:** 2025-05-02T09:08:07Z **Abstract:** This article introduces Perception-Informed Neural Networks (PrINNs), a framework designed to incorporate perception-based information into neural networks, addressing both systems with known and unknown physics laws or differential equations. Moreover, PrINNs extend the concept of Physics-Informed Neural Networks (PINNs) and their variants, offering a platform for the...Votes: 0GitHub stars: 3
- Physics Informed Llm Pid TuningA physics-informed framework that uses Large Language Model agents for PID tuning of chemical processes, combining closed-loop response features, control-engineering diagnoses, and physics-informed reinforcement learning.Votes: 0GitHub stars: 3
- Physics Infused Video GenerationPhysics-Infused Video Generation via joint modeling of visual and latent physical dynamics. Phantom integrates physical-aware video representation into generation process for physically consistent video synthesis. Use for: physics-aware video generation, physical dynamics modeling, consistent video synthesis, latent physics inference. Activation: Phantom, physics-infused video, physical dynamics, video generation, latent physics, 物理注入视频生成.Votes: 0GitHub stars: 3
- Physics Transformer Pde Function ProjectionPhysics Transformer methodology for PDE prediction using function-projection-based tokenization. Treats physical fields as continuous functions with adaptive local basis functions and locality-preserving spatial patches.Votes: 0GitHub stars: 3
- Physicsguided Foundation Model For Scientific Discovery An Application To Aquatic Science**arXiv ID:** 2502.06084 **Authors:** Runlong Yu, Chonghao Qiu, Robert Ladwig, Paul Hanson, Yiqun Xie, Xiaowei Jia **Published:** 2025-02-10T00:48:10Z **Abstract:** Physics-guided machine learning (PGML) has become a prevalent approach in studying scientific systems due to its ability to integrate scientific theories for enhancing machine learning (ML) models. However, most PGML approaches are tailored to isolated and relatively simple tasks, which limits their applicability to complex system...Votes: 0GitHub stars: 3
- Planning With Transformers Chain Of Computation AnDerived from arXiv:2607.17710 - Planning with Transformers: Chain of Computation and Structured Context WindowsVotes: 0GitHub stars: 3
- Portbench Llm Portfolio BenchmarkPortBench: correlation-aware full-pipeline benchmark for LLM-driven portfolio management. Dual-layer evaluation (static QA + dynamic allocation pipeline) with CEPS metric for compounding reasoning errors. 90% of LLMs fail to beat equal-weight. arXiv:2605.27887Votes: 0GitHub stars: 3
- Prediction Markets Option ArbitragePrediction market pricing benchmark methodology — comparing prediction market prices (Polymarket) with option-implied risk-neutral probabilities from centralized exchanges (Binance/Deribit). Use when analyzing prediction market efficiency, cross-venue price discovery, crypto derivatives pricing, market fragmentation effects, and speculative demand wedges.Votes: 0GitHub stars: 3
- Promptbreeder Selfreferential Selfimprovement Via Prompt Evolution**arXiv ID:** 2309.16797 **Authors:** Chrisantha Fernando, Dylan Banarse, Henryk Michalewski, Simon Osindero, Tim Rocktäschel **Published:** 2023-09-28T19:01:07Z **Abstract:** Popular prompt strategies like Chain-of-Thought Prompting can dramatically improve the reasoning abilities of Large Language Models (LLMs) in various domains. However, such hand-crafted prompt-strategies are often sub-optimal. In this paper, we present Promptbreeder, a general-purpose self-referential self-improvement m...Votes: 0GitHub stars: 3
- Qiskit Llm Code MigrationLLM+RAG methodology for automated Qiskit code migration across versions. Uses taxonomy-based RAG to reduce hallucinations and improve code reliability in Quantum Software Engineering (QSE). Activation: qiskit migration, quantum code migration, QDK version upgrade, quantum API refactoring, quantum software engineering, quantum development kit, quantum code maintenance, LLM quantum code, RAG quantum migration, Qiskit version, quantum technical debtVotes: 0GitHub stars: 3
- Quasirecurrent Neural Networks**arXiv ID:** 1611.01576 **Authors:** James Bradbury, Stephen Merity, Caiming Xiong, Richard Socher **Published:** 2016-11-05T00:31:25Z **Abstract:** Recurrent neural networks are a powerful tool for modeling sequential data, but the dependence of each timestep's computation on the previous timestep's output limits parallelism and makes RNNs unwieldy for very long sequences. We introduce quasi-recurrent neural networks (QRNNs), an approach to neural sequence modeling that alternates convoluti...Votes: 0GitHub stars: 3
- R3 Advertisement Compliance Rectification Via Group Relative ExperienceRigorous content moderation is crucial for online advertising but leads to millions of daily rejections. This scale renders manual rectification infeasible, particularly for video advertisements. Howe. Based on arXiv:2607.07318.Votes: 0GitHub stars: 3
- Rag Contextual EnrichmentSkill for AI agent capabilitiesVotes: 0GitHub stars: 3
- Rag Har Towards Cost Efficient Llm Based Human ActRAG-HAR+: Towards Cost-Efficient LLM-Based Human Activity Recognition for Edge DeploymentVotes: 0GitHub stars: 3
- Reasoning Memory Llm**Source:** arXiv:2412.01885Votes: 0GitHub stars: 3
- Remangpt Remanufacturing Automation LlmFramework for transforming remanufacturing automation using large language models to address variability and uncertainty in end-of-life products.Votes: 0GitHub stars: 3
- Residual Gap Aware Transformer MedicalResidual Gap-Aware Transformer methodology for medium-horizon medical prognosis prediction. Anchors predictions at clinical visits, uses only historical data observed at or before anchor, and predicts future clinical score changes (e.g., 24-month CDR-SB change for Alzheimer's). Addresses baseline severity confounding and irregular biomarker observation patterns. Use when: residual gap prediction, anchor-based medical forecasting, Alzheimer's progression prediction, CDR-SB change forecasting, ...Votes: 0GitHub stars: 3
- Reward Driven Llm Agent Workflows Synthesizing PomDerived from arXiv:2607.17038 - Reward-Driven LLM Agent Workflows: Synthesizing POMDP Routing and Self-Correction for Autonomous Decision-MakingVotes: 0GitHub stars: 3
- Reward Driven Llm Agent Workflows Synthesizing Pomdp Routing Self Correction Autonomous DecSkill derived from arXiv:2607.17038 - Reward-Driven LLM Agent Workflows: Synthesizing POMDP Routing and Self-Correction for Autonomous DecVotes: 0GitHub stars: 3
- Rg Dnn InterpretabilityRenormalization Group (RG) interpretability framework for deep neural networks. Establishes correspondence between RG in statistical physics and DNN training, proving DNN feature extraction is equivalent to RG flow on exponential family distributions. Based on arXiv:2606.00157 (Gong & Xia, 2026).Votes: 0GitHub stars: 3
- Rgflow Transformer EegRG-Flow Transformer for encoding scale-free dynamics in scarce EEG data - uses renormalization-group inductive bias to improve interpretability and spectral exponent recovery from limited neural recordingsVotes: 0GitHub stars: 3
- Rl Post Training Builds Compositional Reasoning StrategiesDoes RL post-training merely amplify primitive skills already latent in a base model, or can it compose primitive skills into new higher-level strategies? We study this question in a fully observable. Based on arXiv:2607.07646.Votes: 0GitHub stars: 3
- Rlhf Helpful Harmless AssistantAnthropic's RLHF training methodology for creating helpful and harmless AI assistants using preference modeling and reinforcement learning from human feedbackVotes: 0GitHub stars: 3
- Role Based Llm FrameworkImplement role-based multi-agent LLM frameworks for complex domain tasks. Use when extracting structured information from diverse documents, building domain-specific agent teams, or reducing hallucinations through role specialization. Based on arXiv:2604.01529 - A Role-Based LLM Framework for Structured Information Extraction from Healthy Food Policies.Votes: 0GitHub stars: 3
- San Hypothesizing Longterm Synaptic Development And Neural Engram Mechanism In Scalable Models Parameterefficient Finetuning**arXiv ID:** 2409.06706 **Authors:** Gaole Dai, Chun-Kai Fan, Yiming Tang, Zhi Zhang, Yuan Zhang, Yulu Gan, Qizhe Zhang, Cheng-Ching Tseng, Shanghang Zhang, Tiejun Huang **Published:** 2024-08-24T03:27:29Z **Abstract:** Advances in Parameter-Efficient Fine-Tuning (PEFT) bridged the performance gap with Full Fine-Tuning (FFT) through sophisticated analysis of pre-trained parameter spaces. Starting from drawing insights from Neural Engrams (NE) in Biological Neural Networks (BNNs), we establis...Votes: 0GitHub stars: 3
- Scaling Laws For Neural Language ModelsSkill for AI agent capabilitiesVotes: 0GitHub stars: 3
- Score Llm Self CorrectionSkill for AI agent capabilitiesVotes: 0GitHub stars: 3
- Search Fail Recover A Training Framework For Correction Aware ReasoningMany reasoning tasks are not well described by a single left-to-right chain: a solver may need to pursue a plausible branch, observe delayed failure, and return to the latest prefix that can still be. Based on arXiv:2607.07492.Votes: 0GitHub stars: 3
- Search Hardness Aware Llm Problem FormulationSearch Hardness-Aware LLM-Based Problem Formulation (SHA-PF) framework for expensive simulation-driven design. Prioritizes formulations that guide efficient search by focusing on rare samples with greater progress potential, reducing evaluation requirements significantly.Votes: 0GitHub stars: 3
- Secure Coding Drift Llm PqcGamified methodology for detecting and fixing secure coding drift in LLM-assisted post-quantum cryptography development. Identifies gradual degradation of secure coding practices from sustained reliance on LLM-generated code in security-critical domains.Votes: 0GitHub stars: 3
- Selectinfer Selective Neuron Loading Computation Device LlmsSkill derived from arXiv:2607.18081 - SelectInfer: Selective Neuron Loading and Computation for On-Device LLMsVotes: 0GitHub stars: 3
- Serf Towards Better Training Of Deep Neural Networks Using Logsoftplus Error Activation Function**arXiv ID:** 2108.09598 **Authors:** Sayan Nag, Mayukh Bhattacharyya **Published:** 2021-08-21T23:33:57Z **Abstract:** Activation functions play a pivotal role in determining the training dynamics and neural network performance. The widely adopted activation function ReLU despite being simple and effective has few disadvantages including the Dying ReLU problem. In order to tackle such problems, we propose a novel activation function called Serf which is self-regularized and nonmonotonic in n...Votes: 0GitHub stars: 3
- Serverlesst2i Efficient Text To Image Workflow SerServerlessT2I: Efficient Text-to-Image Workflow Serving on a Serverless PlatformVotes: 0GitHub stars: 3
- Show Me Your Nft And I Tell You How It Will Perform Multimodal Representation Learning For Nft Selling Price Prediction**arXiv ID:** 2302.01676 **Authors:** Davide Costa, Lucio La Cava, Andrea Tagarelli **Published:** 2023-02-03T11:56:38Z **Abstract:** Non-Fungible Tokens (NFTs) represent deeds of ownership, based on blockchain technologies and smart contracts, of unique crypto assets on digital art forms (e.g., artworks or collectibles). In the spotlight after skyrocketing in 2021, NFTs have attracted the attention of crypto enthusiasts and investors intent on placing promising investments in this profitable...Votes: 0GitHub stars: 3
- Simple Tokenizer Agnostic On Policy DistillationSimpleOPD for cross-tokenizer on-policy distillation.Votes: 0GitHub stars: 3
- Skill Rag IndexerRAG indexer for local skill documents with semantic search and intelligent skill recommendation.Votes: 0GitHub stars: 3
- Smartvector Self Aware Embedding RagSmartVector self-aware vector embedding framework for RAG inspired by neuroscience. Hippocampus-neocortex memory consolidation model with four-signal retrieval (temporal, confidence, relational, contextual). Activation: RAG, vector embedding, self-aware embedding, hippocampus, neocortex, memory consolidation, temporal embedding, confidence weighting, knowledge retrieval.Votes: 0GitHub stars: 3
- Smetric Llm Scheduling Serving Agents Session CentricBalanced session-centric LLM scheduling for agent serving workloads. Routes first request in each session for load balance and follow-ups cache-aware. 10-16% TPS improvement. Leverages intra-session locality and 80%+ KV-reuse in agent traces. Activation: LLM scheduling, agent serving, session-centric scheduling, inference infrastructure, tokens-per-second.Votes: 0GitHub stars: 3
- Snuffy Efficient Whole Slide Image Classifier**arXiv ID:** 2408.08258 **Authors:** Hossein Jafarinia, Alireza Alipanah, Danial Hamdi, Saeed Razavi, Nahal Mirzaie, Mohammad Hossein Rohban **Published:** 2024-08-15T16:59:15Z **Abstract:** Whole Slide Image (WSI) classification with multiple instance learning (MIL) in digital pathology faces significant computational challenges. Current methods mostly rely on extensive self-supervised learning (SSL) for satisfactory performance, requiring long training periods and considerable computationa...Votes: 0GitHub stars: 3
- Sok Security Autonomous Llm AgentsResearch paper: SoK - Security of Autonomous LLM Agents in Agentic Commerce. Comprehensive security analysis of autonomous agents in commercial settings.Votes: 0GitHub stars: 3
- Spacer Towards Engineered Scientific Inspiration**arXiv ID:** 2508.17661 **Authors:** Minhyeong Lee, Suyoung Hwang, Seunghyun Moon, Geonho Nah, Donghyun Koh, Youngjun Cho, Johyun Park, Hojin Yoo, Jiho Park, Haneul Choi, Sungbin Moon, Taehoon Hwang, Seungwon Kim, Jaeyeong Kim, Seongjun Kim, Juneau Jung **Published:** 2025-08-25T04:49:16Z **Abstract:** Recent advances in LLMs have made automated scientific research the next frontline in the path to artificial superintelligence. However, these systems are bound either to tasks of narrow scope...Votes: 0GitHub stars: 3
- Sparse Delta Memory Scaling The State Of Linear Rnns Through SparsityLinear attention models allow a fixed state size and a fixed amount of compute per token. However, due to their limited state size, linear attention models fall behind in long-context recall compared. Based on arXiv:2607.07386.Votes: 0GitHub stars: 3
- Sparse Linear Expert TransformerSparsely gated tiny linear experts (sgatlin) methodology — replacing transformer feedforward layers with networks of sparsely-gated linear neurons for improved compute efficiency and interpretability. Covers isoflop comparison, linear expert removal of nonlinearity, semantic cluster interpretation, and causal factual recall. Activation: sparse linear expert, sgatlin, tiny linear expert, sparsely gated linear, linear MoE, sparse transformer FFN.Votes: 0GitHub stars: 3
- Ssm Contraction ControlController design for Structured State-space Models (SSMs) using contraction theory with indirect data-driven output feedback. Use when: (1) designing controllers for nonlinear systems identified via SSM surrogate models, (2) implementing contraction-based stabilization with Linear Matrix Inequality (LMI) conditions, (3) establishing separation principle for observer-controller design, (4) applying scalable control design to time-series and dynamical systems. First controllability/observabili...Votes: 0GitHub stars: 3
- Stochastic Neural Networks For Hierarchical Reinforcement Learning**arXiv ID:** 1704.03012 **Authors:** Carlos Florensa, Yan Duan, Pieter Abbeel **Published:** 2017-04-10T18:41:28Z **Abstract:** Deep reinforcement learning has achieved many impressive results in recent years. However, tasks with sparse rewards or long horizons continue to pose significant challenges. To tackle these important problems, we propose a general framework that first learns useful skills in a pre-training environment, and then leverages the acquired skills for learning faster in d...Votes: 0GitHub stars: 3