Machine learning news — deep learning, reinforcement learning, neural architecture search, diffusion models, and new ML frameworks and libraries.
An engineer at Frontend Facile explains the Transformer architecture, which underpins modern AI models like GPT, Gemini, and Claude. The Transformer overcomes two key limitations of previous recurrent architectures: poor…
The ELLIS Society approved its fourth summer school on Machine Learning for Healthcare and Biology, following three successful events between 2023 and 2025. The three-day event featured 12 expert researchers from 8 Europ…
A new study, 'The Physics of Multi-Turn Long-Horizon Planning,' introduces a controlled environment to isolate variables in agentic planning. The research identifies three stages—pre-training, post-training, and integrat…
A developer demonstrated that prompt-to-model pipelines for computer vision are becoming feasible on consumer hardware by generating nearly 1,000 synthetic images of wrenches using Flux Klein 9B, annotating them with Met…
A product manager developing a business matching system using similarity search is advised to adopt retrieval-augmented generation (RAG) and combine embeddings with structured metadata such as industry, location, and ser…
A Hugging Face user asks for the best method to extract services from free text, citing a rule-based approach with synonym dictionaries and negation detection as a starting point, with NLP or zero-shot models as a potent…
Researchers trained a multi-label classifier on a curated gold dataset of 1,207 CVEs from expert MITRE Center for Threat-Informed Defense mappings, achieving recall@5 of 0.673 ± 0.019 for mapping CVEs to MITRE ATT&CK tec…
A developer built a reproducible sales-forecasting workflow for a B2B SaaS company that combines CRM extraction, stage calibration, machine-learning win probabilities, and Power BI dashboards. The system uses determinist…
A new study from researchers including Elisaweta Masserova reveals that cryptographic model certification (CMC) schemes built on zero knowledge proofs can be attacked by model providers who engineer training data, allowi…
Six Android malware detectors in wide research use, including Drebin, MalScan, MaskDroid, and the LLM-based LAMD, flagged more than half the benign apps in a test set assembled from 49 Google Play categories, according t…
A new study introduces CaRE, a compute-aware evaluation framework for masked diffusion language models (MDLMs), revealing that current evaluation standards conflate algorithmic improvements with hidden choices of compute…
Researchers propose SymCA, an LLM-empowered interpretable column annotation framework that materializes column annotation as a global-to-local symbolic decision process, outperforming the strongest baselines by an averag…
Researchers at HKBU KnowComp introduce VisualPatchWorld (VPW), a method that represents world dynamics as code for planning, achieving 69.0% mean planning success and outperforming the strongest code baseline by 23.5 poi…
A new study from arXiv (arXiv:2607.25270v1) finds that activation steering in language models is substantially affected by the choice of source activations, with effective steering signals coming from execution-boundary …
TabRank, a framework for training reasoning rerankers for tabular retrieval, improves Acc@10 by 30.5% on HybridQA, 15.2% on SQA, 52.9% on TabFact, and 13.1% on TATQA compared to the base model, according to a new arXiv p…
A new arXiv preprint (2607.25202v1) presents a cross-lingual analysis of conversational entrainment in Mandarin-English, Hindi-English, and Spanish-English code-switched speech, finding that lexical entrainment generaliz…
Researchers propose temporal-distance JEPA (TD-JEPA), a plan-aware representation learning method that mines directed temporal cost from reward-free trajectories to improve latent world model predictive control. Under lo…
Researchers introduce Generative Topology Optimization (GenTO), a unified framework that trains a diffusion model on a large full-order topology dataset and iteratively steers the resulting topology distribution toward t…
A deep neural model combining multi-layer temporal convolution networks with label-wise attention improves medical coding accuracy, achieving a 9% increase in F-1 scores and a 28% increase in recall over the previous sta…
Researchers propose WHTMix, a Walsh-Hadamard token mixer that replaces global self-attention in stereo depth estimation transformers, reducing model compute by a factor of 2.46 and single-image inference latency by a fac…