Machine learning news — deep learning, reinforcement learning, neural architecture search, diffusion models, and new ML frameworks and libraries.
Sapient Inc. open-sourced HRM-Text in May 2026, a 1.15 billion parameter text generation model based on the HRM architecture. Trained on roughly 40 billion tokens—up to 1,000 times less data than comparable models—it ach…
A tech writer fine-tuned a local LLM to replicate the style of 1980s and 1990s software documentation, using a 37-million-word corpus of scanned Microsoft manuals from the Bitsavers archive. The experiment, conducted on …
A new raw-waveform diffusion model called WavFlow has achieved audio fidelity matching or exceeding that of autoencoder-based pipelines, eliminating the need for latent compression. On the VGGSound benchmark, WavFlow's F…
Megan Lawther is developing Guardian AI, a forensic diagnostic decision-support system designed to improve clarity in complex clinical pathology. The system uses Google's MedGemma, TensorFlow, and Edge AI to cross-refere…
Google released Magenta RealTime 2, an open-weights AI model and real-time inference engine that enables live, low-latency music generation on a MacBook. The model responds instantly to MIDI, audio, and text inputs, achi…
Researchers have developed a deep learning algorithm that restores three-dimensional capillary anatomy from a single optical coherence tomographic angiography (OCTA) volume, addressing imaging artifacts that hinder relia…
Researchers have developed BUMP, a self-supervised framework that trains large language models to generate personalized user profiles without requiring labeled data from downstream tasks. The system uses a bidirectional …
Researchers have developed NIV (Neural Axis Variations), a method that automatically converts a static font into a fully functional variable font by predicting per-point displacements along desired design axes. Trained o…
Researchers have developed a cross-model safety steering framework that transfers a safety direction from a source large language model to a target image or video generator using only benign data, eliminating the need fo…
Researchers have developed Agentic Monte Carlo (AMC), a method that enables reinforcement learning-style optimization of black-box LLM agents without requiring parameter-level access. The technique uses Sequential Monte …
Researchers at a major AI lab have identified a specific neural subgraph within a large language model (Qwen3-4B-Instruct-2507) that governs how the model trades off short-term gains against long-term consequences. The t…
Researchers introduced Branched Neural Rough Differential Equations (B-NRDEs), a Hopf-algebraic framework that extends neural rough differential equations to handle Itô dynamics and manifold-valued data. The method uses …
Researchers have developed PRECISE, a method that combines small human-labeled datasets with large LLM-generated judgments to produce bias-corrected estimates of ranking evaluation metrics. The approach reduces standard …
Researchers have developed ReGeN, a reference-guided generative pipeline that creates synthetic multivariate time series data by decomposing observed sequences into periodic backbones, stochastic residuals, and cross-var…
Researchers have developed CVT-RL, a reinforcement learning algorithm that uses policy-conditioned counterfactual credit assignment to reduce unsupported evidence chains and shortcut actions in long-horizon language agen…
A study of behavioral foundation models trained on user event sequences found that a small embedder using only 2% of parameters is compute-optimal across all tested budgets, as embedder parameters are more expensive per …
Researchers introduced Flash-WAM, a modality-aware distillation framework that compresses world-action model inference to a single step, reducing per-chunk latency from 8.1 seconds to 348 milliseconds on NVIDIA L40S hard…
Researchers have developed DiffSlack, a differentiable projection layer that enforces nonlinear inequality constraints in neural networks by reformulating them as equalities with learnable slack variables. Tested on vehi…
Researchers have introduced state commitment learning, a new training objective that teaches language models to distinguish temporary computation from persistent memory. The method, called Counterfactual Erasure RL (CERL…
Researchers have developed ReasoningFlow, a framework that maps the non-linear discourse structures of large reasoning model (LRM) traces into directed acyclic graphs (DAGs) to improve evaluation and monitoring. Through …