AI Infrastructure

GPUs, training clusters, MLOps, and deployment

This Week's Best | Monthly Best | Guide | Trending

Top This Week

Ai Infrastructure

Most people are using AI wrong—and it’s capping what they can do

1 is a fluke. 2 is a coincidence. 3 is a pattern. Lately I’ve been noticing something. The problems I’m solving are getting more complex…...

Reddit - Artificial Intelligence · 1 min ·
Ai Infrastructure

Most people are using AI wrong—and it’s capping what they can do

1 is a fluke. 2 is a coincidence. 3 is a pattern. Lately I’ve been noticing something. The problems I’m solving are getting more complex…...

Reddit - Artificial Intelligence · 1 min ·
Machine Learning

[P] ML project (XGBoost + Databricks + MLflow) — how to talk about “production issues” in interviews?

Hey all, I recently built an end-to-end fraud detection project using a large banking dataset: Trained an XGBoost model Used Databricks f...

Reddit - Machine Learning · 1 min ·

All Content

[2506.21220] Complexity-aware fine-tuning
Llms

[2506.21220] Complexity-aware fine-tuning

The paper presents a novel method for fine-tuning large language models (LLMs) by categorizing training data based on complexity, resulti...

arXiv - Machine Learning · 3 min ·
[2506.10914] Foundation Models for Causal Inference via Prior-Data Fitted Networks
Llms

[2506.10914] Foundation Models for Causal Inference via Prior-Data Fitted Networks

This paper introduces CausalFM, a framework for training prior-data fitted networks (PFNs) for causal inference, enhancing Bayesian infer...

arXiv - Machine Learning · 4 min ·
[2503.12016] A Survey on Federated Fine-tuning of Large Language Models
Llms

[2503.12016] A Survey on Federated Fine-tuning of Large Language Models

This survey explores the integration of Federated Learning with Large Language Models (LLMs), addressing challenges and methodologies for...

arXiv - Machine Learning · 4 min ·
[2501.08219] Characterizing LLM Inference Energy-Performance Tradeoffs across Workloads and GPU Scaling
Llms

[2501.08219] Characterizing LLM Inference Energy-Performance Tradeoffs across Workloads and GPU Scaling

The paper explores the energy-performance tradeoffs in LLM inference across various workloads and GPU scaling, revealing significant insi...

arXiv - Machine Learning · 4 min ·
[2602.21144] Scaling State-Space Models on Multiple GPUs with Tensor Parallelism
Llms

[2602.21144] Scaling State-Space Models on Multiple GPUs with Tensor Parallelism

This paper explores tensor parallelism for scaling selective state-space models (SSMs) on multiple GPUs, addressing challenges in memory ...

arXiv - Machine Learning · 4 min ·
[2602.21039] Is Multi-Distribution Learning as Easy as PAC Learning: Sharp Rates with Bounded Label Noise
Ai Infrastructure

[2602.21039] Is Multi-Distribution Learning as Easy as PAC Learning: Sharp Rates with Bounded Label Noise

This paper investigates the complexities of multi-distribution learning, revealing that achieving fast learning rates is inherently more ...

arXiv - Machine Learning · 4 min ·
[2602.20816] Don't Ignore the Tail: Decoupling top-K Probabilities for Efficient Language Model Distillation
Llms

[2602.20816] Don't Ignore the Tail: Decoupling top-K Probabilities for Efficient Language Model Distillation

The paper presents a novel approach to language model distillation by introducing a tail-aware divergence that enhances the influence of ...

arXiv - Machine Learning · 3 min ·
[2602.20712] F10.7 Index Prediction: A Multiscale Decomposition Strategy with Wavelet Transform for Performance Optimization
Machine Learning

[2602.20712] F10.7 Index Prediction: A Multiscale Decomposition Strategy with Wavelet Transform for Performance Optimization

This article presents a novel forecasting method for the F10.7 solar index using wavelet decomposition, demonstrating improved prediction...

arXiv - Machine Learning · 4 min ·
[2602.20646] On the Convergence of Stochastic Gradient Descent with Perturbed Forward-Backward Passes
Ai Infrastructure

[2602.20646] On the Convergence of Stochastic Gradient Descent with Perturbed Forward-Backward Passes

This paper analyzes the convergence of Stochastic Gradient Descent (SGD) under perturbations in both forward and backward passes, providi...

arXiv - Machine Learning · 3 min ·
[2602.20611] Amortized Bayesian inference for actigraph time sheet data from mobile devices
Machine Learning

[2602.20611] Amortized Bayesian inference for actigraph time sheet data from mobile devices

This article presents a novel approach to Bayesian inference for analyzing actigraph time sheet data from mobile devices, focusing on hea...

arXiv - Machine Learning · 4 min ·
[2602.20450] Heterogeneity-Aware Client Selection Methodology For Efficient Federated Learning
Machine Learning

[2602.20450] Heterogeneity-Aware Client Selection Methodology For Efficient Federated Learning

The paper presents Terraform, a novel client selection methodology for federated learning that addresses client heterogeneity, achieving ...

arXiv - Machine Learning · 3 min ·
[2602.20178] Data-Driven Deep MIMO Detection:Network Architectures and Generalization Analysis
Machine Learning

[2602.20178] Data-Driven Deep MIMO Detection:Network Architectures and Generalization Analysis

This paper presents a data-driven approach to Multiuser Multiple-Input Multiple-Output (MU-MIMO) detection, introducing a novel architect...

arXiv - Machine Learning · 4 min ·
[2602.20164] Benchmarking Distilled Language Models: Performance and Efficiency in Resource-Constrained Settings
Llms

[2602.20164] Benchmarking Distilled Language Models: Performance and Efficiency in Resource-Constrained Settings

This paper benchmarks distilled language models, demonstrating their superior performance and efficiency in resource-constrained environm...

arXiv - Machine Learning · 3 min ·
[2602.21196] Untied Ulysses: Memory-Efficient Context Parallelism via Headwise Chunking
Machine Learning

[2602.21196] Untied Ulysses: Memory-Efficient Context Parallelism via Headwise Chunking

The paper presents UPipe, a novel technique for memory-efficient context parallelism in Transformer models, achieving significant reducti...

arXiv - Machine Learning · 4 min ·
[2602.21168] Sequential Counterfactual Inference for Temporal Clinical Data: Addressing the Time Traveler Dilemma
Machine Learning

[2602.21168] Sequential Counterfactual Inference for Temporal Clinical Data: Addressing the Time Traveler Dilemma

This article presents a novel Sequential Counterfactual Framework for analyzing temporal clinical data, addressing limitations of traditi...

arXiv - Machine Learning · 4 min ·
[2602.21078] ProxyFL: A Proxy-Guided Framework for Federated Semi-Supervised Learning
Machine Learning

[2602.21078] ProxyFL: A Proxy-Guided Framework for Federated Semi-Supervised Learning

The article presents ProxyFL, a novel framework for Federated Semi-Supervised Learning (FSSL) that addresses data heterogeneity issues by...

arXiv - Machine Learning · 4 min ·
[2602.21081] Scaling Vision Transformers: Evaluating DeepSpeed for Image-Centric Workloads
Llms

[2602.21081] Scaling Vision Transformers: Evaluating DeepSpeed for Image-Centric Workloads

This article evaluates the use of DeepSpeed to enhance the scalability of Vision Transformers (ViTs) for image-centric workloads, focusin...

arXiv - Machine Learning · 3 min ·
[2602.20937] Extending $μ$P: Spectral Conditions for Feature Learning Across Optimizers
Llms

[2602.20937] Extending $μ$P: Spectral Conditions for Feature Learning Across Optimizers

This article presents a framework for extending the maximal update parameterization ($μ$P) to various optimizers, enhancing feature learn...

arXiv - Machine Learning · 4 min ·
[2602.20593] Is the Trigger Essential? A Feature-Based Triggerless Backdoor Attack in Vertical Federated Learning
Machine Learning

[2602.20593] Is the Trigger Essential? A Feature-Based Triggerless Backdoor Attack in Vertical Federated Learning

This paper presents a novel feature-based triggerless backdoor attack in vertical federated learning, demonstrating that triggers are not...

arXiv - Machine Learning · 4 min ·
[2602.20427] GauS: Differentiable Scheduling Optimization via Gaussian Reparameterization
Machine Learning

[2602.20427] GauS: Differentiable Scheduling Optimization via Gaussian Reparameterization

The paper presents GauS, a novel differentiable framework for operator scheduling that utilizes Gaussian distributions to optimize schedu...

arXiv - Machine Learning · 3 min ·
Previous Page 82 Next

Related Topics

Machine Learning Large Language Models Generative AI Natural Language Processing Computer Vision Robotics & Embodied AI

Stay updated with AI News

Get the latest news, tools, and insights delivered to your inbox.

Subscribe to Newsletter

Daily or weekly digest • Unsubscribe anytime