Machine learning news — deep learning, reinforcement learning, neural architecture search, diffusion models, and new ML frameworks and libraries.
A new method called SlideRuler calibrates pathology foundation models from a single slide scan, reducing mean target-to-source embedding distance by 16.3-38.5% relative to raw embeddings across two encoders and five SCOR…
Researchers introduced S2Tok, a feed-forward framework that maintains a size-adaptive, persistent scene state from uncalibrated image streams to enable streaming 3D Gaussian reconstruction, according to arXiv paper 2610.…
Researchers introduced DISRQAD, a subjective-quality dataset and diagnostic benchmark containing mean opinion scores for 14,000 super-resolution outputs from ten diffusion and four non-diffusion methods across four low-r…
A new arXiv paper (2610.09031v1) introduces a plug-and-play framework for concept-based interpretation and refinement of medical imaging models, building a Concept Bottleneck Model by aligning a single-modality encoder t…
Researchers introduced GUARD, a geometric uncertainty-aware framework that performs point filtering and segmentation in a single forward pass for robotic hard disk drive (HDD) disassembly, improving PointNet++ segmentati…
Microsoft researchers released Anaximander, an open-source system that runs geospatial deep learning models on any compute backend behind one interactive interface, with code at github.com/microsoft/nxmndr. The system pa…
LayerRoPE, a depth-wise positional encoding method presented in arXiv paper 2610.09179v1, replaces all layerwise normalization weight vectors in a Transformer with a single shared vector and depth-conditioned scalars, cu…
A new arXiv paper (2610.09108v1) shows that the exact per-iteration policy-learning objective in reinforcement learning, written in log-density-ratio coordinates y := log[π/π_n] and computed via per-decision importance s…
A new arXiv paper (arXiv:2610.09174v1) introduces the Context-Aware Attention-based Gaussian Mixture Model (CAA-GMM), a multimodal, uncertainty-aware framework for vehicular trajectory prediction that models future motio…
A new arXiv paper (2610.09186v1) argues that step-by-step reasoning and pattern recognition in large language models lie on a spectrum defined by how much preceding context each next token depends on, and proves the numb…
A comparative study of multi-label topic assignment for user-generated content found that while discriminative models (DeBERTa-V3 and ModernBERT) outperform ultra-lightweight generative models on structured product revie…
Researchers introduced MaRK (Markov-adapted Recurrent Kernels), a dynamic operator-conditioning framework that maps context vectors directly into bounded modulations of a frozen State Space Model's recurrence (A), read-i…
A new arXiv paper (2610.09120v1) shows that an adversarial agent can exploit leaked signals merely correlated with a victim agent's randomized exploration to steer multi-agent bandit learning toward a new steady state th…
Researchers introduced Q-PACE, a dynamic precision allocation method for quantization-aware training that uses a second-order sensitivity model to predict loss increases as quantization noise MSE weighted by per-layer cu…
A new arXiv paper, arXiv:2610.09196v1, extends the known connection between linear recursive feature machines (RFMs) and iteratively reweighted least squares from the interpolating setting to ridge-regularized multi-outp…
A paper on arXiv (2610.09145v1) reports that noising conditioning prompt tokens during training, rather than keeping them clean, improves continuous diffusion language models on combinatorial reasoning tasks. The single-…
A new arXiv paper (2610.09250v1) proposes a sample-only framework for evaluating and selecting Best-of-N (BoN) inference-time alignment policies without access to response likelihoods, which standard off-policy estimator…
Researchers introduced LeCuration, a small world model built on the LeWorldModel (LeWM) latent encoder and predictor with an added diffusion transformer (DiT) decoder, to curate data for physical AI applications. In a qu…
Researchers introduced N-GNAS, a Node-Level Graph Neural Architecture Search algorithm that automatically selects an appropriate network architecture for each subset of nodes when updating node features, according to an …
A new arXiv paper (2610.09247v1) identifies an "informational curse of horizon" in goal-conditioned policy learning, where increasing the goal relabeling horizon significantly reduces policy generalization and performanc…