Machine learning news — deep learning, reinforcement learning, neural architecture search, diffusion models, and new ML frameworks and libraries.
The MiMo-V2.5 model family introduces Hybrid Sliding Window Attention (Hybrid SWA) to reduce KVCache storage to roughly 1/7 of Full Attention, sparse MoE activation to cut per-token compute, and multimodal encoders for c…
Ternlight, a 7 MB embedding model using ternary quantization and WebAssembly, enables fully client-side semantic search in browsers, bypassing server round-trips and API costs. The model runs locally via a lightweight WA…
WebAssembly (WASM) and the Ternlight framework enable compact, high-performance AI deployment for low-bandwidth areas. The approach uses model compression, WASM conversion, and offline operation to deliver AI on resource…
Jim Dowling's O'Reilly book 'Building Machine Learning Systems with a Feature Store' offers a practical guide for deploying machine learning models into production, based on a course he taught at KTH in Stockholm. The bo…
A data engineer outlines a practical roadmap for breaking into the field in 2026, emphasizing SQL fluency, Python scripting, cloud cost optimization, and data modeling over outdated tools like Hadoop. The guide prioritiz…
Scientists at Aalto University and international collaborators combined machine learning with quantum physics to discover two new superconductors, YRu3B2 and LuRu3B2, and developed a faster method to search for room-temp…
A developer evaluated the YAMNet TFLite model on the ESC-50 environmental sound dataset, achieving 60.45% top-1 accuracy across 2,000 clips. The experiment used a pretrained model without fine-tuning, with higher accurac…
Researchers at the Mechanistic Interpretability project have identified that the knight-fork policy logit in the Maia 3 chess transformer snaps into place after block 5's attention layer, using logit lens and attention p…
Researchers introduced SkillOpt-Lite, a minimal skill optimization pipeline for autonomous agents that accelerates convergence and improves performance, achieving +8.8 points on LiveMath with GPT-5.5 and +25.4 points wit…
Researchers introduced QuantFlow, a federated Mamba-based foundation model for time-series forecasting that combines inverted sequence embedding, bidirectional state-space decoders, quantile regression, and federated lea…
Researchers at a university found that selectively applying monocular depth priors from Depth Anything V2 improves sparse-view neural reconstruction in driving scenes, with Splatfacto gaining 1.029 PSNR and reducing RMSE…
Researchers used interpretable machine learning to predict Parkinson's disease motor severity from QSM and fMRI features, achieving 45.4% variance explained and 75% of predictions within 5 points of clinical scores. The …
Researchers propose a sliding-window-based reinforcement learning (SWRL) framework for dynamic assembly flow shop scheduling with multi-product delivery, addressing challenges from dynamic order arrivals and kitting cons…
Researchers introduced VulcanVoxel, a spatial affordance model that learns 3D blade insertion feasibility in cluttered stowing environments using a masked autoencoder over occupancy fields. Trained on 10,000 real warehou…
Researchers propose Object-Centric Environment Modeling (OCM), a method that organizes LLM agent experience into executable object-centric code bases, improving performance and reducing invalid actions across benchmarks.
Researchers developed a two-stage U-Net framework for efficient pedestrian-level wind prediction across urban morphologies, using the UrbanTALES dataset. The first stage predicts wind fields patch-by-patch, while the sec…
Researchers propose a reinforcement learning with verifiable rewards (RLVR) method for data-efficient adaptation of audio-language models to code-switched automatic speech recognition (ASR). Using Qwen2-Audio across 10 l…
Researchers proposed DMKGC, a conditional diffusion-guided knowledge transfer framework for multi-domain knowledge graph completion, achieving a 4.3% average MRR improvement over state-of-the-art methods. The framework t…
Researchers propose TR-RAG, a teacher-regularized reinforcement learning method for cross-lingual retrieval-augmented generation that addresses language drift and unreliable evidence use. The method combines reward optim…
Researchers developed a multilingual, multimodal NLP framework for early detection of fake news and violence-driven mob activity, achieving 98% accuracy on a fused dataset of 138,256 Bangla and English samples. The syste…