Page 115 of 130 • 1560 Total Articles

createLiveAI

Continue exploring the latest AI breakthroughs, technology insights, and industry analysis. Page 115 of our comprehensive AI news collection.

All Articles 1560 Business 249 Ethics 150 General 142 Policy 12 Research 793 Startups 13 Technology 201

📰 Latest Intelligence

Showing 12 articles on page 115 of 130

Live feed

📱 2-column layout

arXiv Machine Learning

Research

📄 arXiv Machine Learning

May 31, 2025

Score-based Generative Modeling for Conditional Independence Testing

A new method for testing conditional independence (CI) in high-dimensional data is proposed, utilizing score-based generative modeling with sliced conditional score matching and Langevin dynamics, which ensures accurate Type I error control and improved power. Extensive experiments demonstrate that this approach outperforms existing GAN-based methods, offering a more reliable and interpretable solution for CI testing in machine learning and statistics.

research machine-learning

arXiv Machine Learning

Research

📄 arXiv Machine Learning

May 31, 2025

SCORPIO: Serving the Right Requests at the Right Time for Heterogeneous SLOs in LLM Inference

The paper presents SCORPIO, an SLO-focused LLM serving system that enhances system goodput and adherence to service level objectives like TTFT and TPOT by leveraging adaptive scheduling and predictive modules. Evaluations show SCORPIO achieves up to 14.4 times higher goodput and 46.5% better SLO compliance compared to existing systems.

arXiv Machine Learning

Research

📄 arXiv Machine Learning

May 31, 2025

SlimLLM: Accurate Structured Pruning for Large Language Models

A new method called SlimLLM is proposed to efficiently prune large language models by evaluating the importance of entire channels and attention heads, enabling better compression with minimal performance loss. The approach, validated on the LLaMA benchmark, outperforms existing methods and achieves state-of-the-art results in structured pruning of LLMs.

Meta AI Transformers

arXiv Machine Learning

Research

📄 arXiv Machine Learning

May 31, 2025

Smart Surrogate Losses for Contextual Stochastic Linear Optimization with Robust Constraints

This research extends contextual stochastic linear optimization to include inequality constraints dependent on uncertain parameters predicted by machine learning models, using conformal prediction-based uncertainty sets. The authors propose the SPO-RC loss and its convex surrogate SPO-RC+ to improve decision-making under constraint uncertainty, demonstrating enhanced performance through experiments and bias correction techniques.

Machine Learning

arXiv Machine Learning

Research

📄 arXiv Machine Learning

May 31, 2025

Test-time augmentation improves efficiency in conformal prediction

A new method using test-time augmentation (TTA) enhances conformal classifiers by reducing their prediction set sizes by 10-14% without retraining models. The approach is flexible, efficient, and effective across multiple datasets, models, and conformal scoring methods, especially under distribution shifts.

research machine-learning

arXiv Machine Learning

Research

📄 arXiv Machine Learning

May 31, 2025

The Panaceas for Improving Low-Rank Decomposition in Communication-Efficient Federated Learning

The paper proposes three innovative techniquesModel Update Decomposition (MUD), Block-wise Kronecker Decomposition (BKD), and Aggregation-Aware Decomposition (AAD)to enhance low-rank decomposition methods in federated learning, addressing key issues of what to decompose, how to decompose, and how to aggregate. Experimental results demonstrate that these methods lead to faster convergence and higher accuracy, supported by a theoretical analysis confirming the convergence of MUD.

arXiv Machine Learning

Research

📄 arXiv Machine Learning

May 31, 2025

Two Is Better Than One: Rotations Scale LoRAs

A new gating method called RadarGate enhances the scalability and generalization of LoRA-based Mixture-of-Experts in large language models by introducing rotational operations among LoRA representations, enabling richer feature interactions. Extensive experiments across multiple benchmarks demonstrate its effectiveness in addressing underfitting and poor generalization issues as the number of LoRAs increases.

research machine-learning

arXiv Machine Learning

Research

📄 arXiv Machine Learning

May 31, 2025

VERINA: Benchmarking Verifiable Code Generation

A new benchmark called Verina has been introduced to evaluate the ability of large language models (LLMs) to generate verifiable code, including code, specifications, and proofs, across 189 curated tasks in Lean. The evaluation reveals significant challenges, with the best model achieving only 61.4% correct code and minimal success in proof generation, highlighting the need for advancements in LLM-based verification methods.

GPT

arXiv Machine Learning

Research

📄 arXiv Machine Learning

May 31, 2025

Walking the Weight Manifold: a Topological Approach to Conditioning Inspired by Neuromodulation

This research introduces a method inspired by neuromodulation, where neural network weights are smoothly parameterized as manifolds conditioned on task variables, enabling better generalization and task transfer. By optimizing these manifolds' topology, such as lines or circles, the approach outperforms traditional input conditioning methods and offers a flexible framework for multi-task learning.

Deep Learning

arXiv Machine Learning

Research

📄 arXiv Machine Learning

May 31, 2025

Weight Spectra Induced Efficient Model Adaptation

This research investigates how fine-tuning large models alters their weight matrices, revealing that it mainly amplifies top singular values and reorients dominant singular vectors, indicating task-specific knowledge is embedded in low-dimensional subspaces. Building on these findings, the authors propose a method that modulates top singular directions through learnable rescaling, improving performance across tasks while maintaining the model's global structure.

arXiv Machine Learning

Research

📄 arXiv Machine Learning

May 31, 2025

When Does Neuroevolution Outcompete Reinforcement Learning in Transfer Learning Tasks?

This paper explores the transfer learning potential of neuroevolution (NE), comparing its performance to reinforcement learning (RL) in tasks requiring skill transfer across increasing complexities. Using new benchmarks, the study finds that NE methods often outperform RL baselines, highlighting NE's promise for developing more adaptable artificial agents, though scaling remains a challenge.

Transformers

arXiv Machine Learning

Research

📄 arXiv Machine Learning

May 31, 2025

X-Factor: Quality Is a Dataset-Intrinsic Property

Research indicates that dataset quality is an intrinsic property, independent of size, class balance, and model architecture, and significantly influences machine learning performance. The study finds that dataset quality emerges from the quality of its constituent classes, making it a key factor alongside size, class balance, and architecture for optimizing classifiers.

Machine Learning

1 2 3 4 5 6 7 ... 130

Page 115 of 130 • Showing articles 1369-1380 of 1560

Quick Navigation

Jump to any page or browse by category

Latest (Page 1) Business 249 Ethics 150 General 142 Policy 12 Research 793 Startups 13