ml stability
30 articles about ml stability in AI news
MLCC Shortage Threatens AI Server Ramp: Prices Hiking, Lead Times Stretching
MLCCs, cheap components stabilizing voltage in AI servers, face supply crunch as demand grows ~5x by CY27. Lead times stretch, prices hike, new lines take 2 years.
Prince Canuma's M3 Ultra 512GB & RTX Pro 6000 Setup for MLX Research
Independent developer Prince Canuma has assembled a powerful, community-sponsored home compute cluster for MLX research and model porting, featuring an M3 Ultra with 512GB RAM and an RTX Pro 6000.
AirTrain Enables Distributed ML Training on MacBooks Over Wi-Fi
Developer @AlexanderCodes_ open-sourced AirTrain, a tool that enables distributed ML training across Apple Silicon MacBooks using Wi-Fi by syncing gradients every 500 steps instead of every step. This makes personal device training feasible for models up to 70B parameters without cloud GPU costs.
MLX Enables Local Grounded Reasoning for Satellite, Security, Robotics AI
Apple's MLX framework is enabling 'local grounded reasoning' for AI applications in satellite imagery, security systems, and robotics, moving complex tasks from the cloud to on-device processing.
mlx-vlm v0.4.2 Adds SAM3, DOTS-MOCR Models and Critical Fixes for Vision-Language Inference on Apple Silicon
mlx-vlm v0.4.2 released with support for Meta's SAM3 segmentation model and DOTS-MOCR document OCR, plus fixes for Qwen3.5, LFM2-VL, and Magistral models. Enables efficient vision-language inference on Apple Silicon via MLX framework.
Qwen3-TTS Added to mlx-tune, Enabling Full Qwen Model Fine-Tuning on Apple Silicon Macs
The mlx-tune library now supports Qwen3-TTS, making the entire Qwen model stack—including the new text-to-speech model—fine-tunable on Apple Silicon Macs. This expands local AI development options for researchers and developers.
MLLMRec-R1: A New Framework for Efficient Multimodal Sequential Recommendation with LLMs
Researchers propose MLLMRec-R1, a framework that makes Group Relative Policy Optimization (GRPO) practical for multimodal sequential recommendation by addressing computational cost and reward inflation issues. This enables more explainable, reasoning-based recommendations.
Fortress Framework Prunes Unstable Features, Boosts Rec Stability by CV
Fortress prunes temporally unstable features in rec models via historical snapshots, improving CV and PR-AUC in offline tests.
Meta, Microsoft Lay Off 17,000 in One Day for AI Spending
Meta fired 8,000 employees and Microsoft laid off 9,000 within hours of each other, signaling a coordinated shift of resources from headcount to AI compute and model development. The layoffs underscore a trend where big tech prioritizes AI investment over workforce stability.
Building a Real-World Fraud Detection System: Beyond Just Training a Model
The article provides a practical breakdown of how to build a production-ready fraud detection system, emphasizing the integration of payment models, sequence models, and shadow mode deployment. It moves beyond pure model training to focus on the operational ML system.
Beyond Dense Connectivity: Explicit Sparsity for Scalable Recommendation
A new arXiv paper introduces SSR, a framework that builds explicit sparsity into recommendation model architectures. It addresses the inefficiency of dense models (like MLPs) when processing high-dimensional, sparse user data, showing superior performance and scalability on datasets including AliExpress.
RLSD Unifies Self-Distillation & Verifiable Rewards to Fix RL Leakage
Researchers propose RLSD, a method merging on-policy self-distillation with verifiable rewards to fix information leakage and training instability in language model reinforcement learning.
How to Fix Claude Code's Remote Control Issues and Get Visual Feedback
Practical solutions for Claude Code's remote control instability and lack of visual feedback when building UI components.
DACT: A New Framework for Drift-Aware Continual Tokenization in Generative Recommender Systems
Researchers propose DACT, a framework to adapt generative recommender systems to evolving user behavior and new items without costly full retraining. It identifies 'drifting' items and selectively updates token sequences, balancing stability with plasticity. This addresses a core operational challenge for real-world, dynamic recommendation engines.
Apple's Neural Engine Jailbroken: Researchers Unlock On-Device AI Training Capabilities
A researcher has reverse-engineered Apple's private Neural Engine APIs to enable direct transformer training on M-series chips, bypassing CoreML restrictions. This breakthrough could enable battery-efficient local model training and fine-tuning without cloud dependency.
Anthropic's 'Harness' Design: What It Means for Your Claude Code Workflows
Anthropic's new 'harness' architecture for long-running apps could enable Claude Code to manage more complex, persistent development tasks with greater stability.
Kimi's Selective Layer Communication Improves Training Efficiency by ~25% with Minimal Inference Overhead
Kimi has developed a method that replaces uniform residual connections with selective information routing between layers in deep AI models. This improves training stability and achieves ~25% better compute efficiency with negligible inference slowdown.
Meituan Open-Sources 1.6T-Parameter LongCat-2.0 Trained on Domestic Chips
Meituan open-sourced 1.6T-parameter LongCat-2.0 trained on 50,000 domestic ASICs, claiming China's first full-process domestic-chip trillion-parameter model.
Cline v4.0.0 Ships Plugin Marketplace
Cline v4.0.0 introduces a plugin marketplace, queued prompts, and an SDK rewrite. Claude Code users get new extensibility and reliability features.
Claude Code Digest — May 14–May 17
Cut CLAUDE.md token waste by 99.3% with progressive disclosure skills.
Moore Threads Q1 Revenue Up, Building 100K-GPU AI Cluster
Moore Threads reports Q1 2026 revenue growth and confirms progress building a 100,000-GPU cluster for AI training, signaling growing domestic AI infrastructure in China despite US export controls.
ASPIRE: New Framework Makes Spectral Graph Filters Learnable for
Researchers propose ASPIRE, a bi-level optimization framework that makes spectral graph filters fully learnable for collaborative filtering, solving the 'low-frequency explosion' problem and matching task-specific designs.
Microsoft's TRELLIS.2: 4B Model Turns Images to 3D in 3 Seconds
Microsoft released TRELLIS.2, a 4B parameter open-source model that generates fully textured, physically accurate 3D models with PBR materials from a single image in about 3 seconds, handling complex geometry like open surfaces and hollow interiors.
TF-LLMER: A New Framework to Fix Optimization Problems in LLM-Enhanced
Researchers identify two key causes of poor training in LLM-enhanced recommenders: norm disparity and misaligned angular clustering. Their solution, TF-LLMER, uses embedding normalization and Rec-PCA to significantly outperform existing methods.
Anthropic Survey: 81,000 People Rank AI Economic Hopes & Fears
Anthropic published new research analyzing the economic hopes and worries expressed by 81,000 people in a prior survey on AI. The findings aim to guide AI development toward public priorities.
CS3: A New Framework to Boost Two-Tower Recommenders Without Slowing Them Down
Researchers propose CS3, a plug-and-play framework that strengthens the ubiquitous two-tower recommendation architecture. It uses three novel mechanisms to improve model alignment and knowledge transfer, delivering significant revenue gains in a live ad system while maintaining millisecond latency.
UALink 2.0 Spec Finalized, Aims to Challenge NVLink for AI Clusters
The UALink 2.0 interconnect specification has been finalized, providing a standardized way to link AI accelerators from AMD, Intel, and others. However, it lags behind NVIDIA's established NVLink technology in real-world deployment.
Quantum Breakthrough: 100,000 Qubits Now Threatens Encryption
The estimated qubits required to break RSA encryption has collapsed from 1 billion in 2012 to just 10,000 in 2026, based on recent papers from Caltech, Google, and quantum startup Oratomic.
Alibaba's DCW Fixes SNR-t Bias in Diffusion Models, Boosts FLUX & EDM
Alibaba researchers developed DCW, a wavelet-based method to correct SNR-t misalignment in diffusion models. The fix improves performance for models like FLUX and EDM with minimal computational cost.
AI-Powered PS4 Emulator 'Spine' Runs Bloodborne Locally on PC
A developer has released Spine, a PS4 emulator that uses AI techniques to run Bloodborne fully on PC. This represents a major step forward in console emulation, previously considered years away.