Explore documented scientific discoveries, original research and new scientific developments. Search the library by topic, category or date.
arXiv:2610.37934v1 Announce Type: new Abstract: Pre-training foundational models is computationally expensive as models scale. In this work, we introduce Bi-Phase Training (BPT), a parameter-efficient pre-training method that preserves the high-rank learning dynamics observed in full-parameter optimization with substantially reduced trainable parameters.
arXiv MLRead article →arXiv:2610.37914v1 Announce Type: new Abstract: Many open mixture-of-experts (MoE) language models train with Muon instead of AdamW. Most balance the load with a bias on each expert's routing score, adjusted after every step. Their recipes disagree on whether Muon should also train the router.
arXiv MLRead article →arXiv:2610.37910v1 Announce Type: new Abstract: Stellarators offer a promising magnetic confinement approach for achieving controlled nuclear fusion. However, stellarator design remains challenging, as evaluating configuration performance requires computationally expensive, high-fidelity simulations.
arXiv MLRead article →arXiv:2610.37909v1 Announce Type: new Abstract: Distributed fire-monitoring systems may involve many sensing devices reporting over a shared uplink, even though the operational objective is often to identify regions exhibiting evidence of fire rather than to recover device-specific measurements.
arXiv MLRead article →arXiv:2610.37903v1 Announce Type: new Abstract: Metis is a bet that better data, better training methods and new architectures can buy what a bigger parameter count cannot, at a fraction of the cost. Metis-1.6 is its latest step: two language models built from scratch by a single researcher, Praxis (2.94B parameters, 0.46B active per pass) and Logos (10.23B parameters, 1.18B active per pass).
arXiv MLRead article →arXiv:2610.37898v1 Announce Type: new Abstract: Graph Contrastive Learning (GCL) has achieved remarkable success in self-supervised representation learning by maximizing agreement between augmented views of a graph. Although augmented view generation is the central step in GCL, existing models rely on stochastic augmentation of the input graph, such as edge perturbation and node masking, which lack structural…
arXiv MLRead article →arXiv:2610.37880v1 Announce Type: new Abstract: Equation learning seeks to discover compact, interpretable equations from data using various machine-learning techniques. However, few workflows discover and interpret learned equations in sparse, high-dimensional materials datasets.
arXiv MLRead article →arXiv:2610.37842v1 Announce Type: new Abstract: As growing compute budgets outpace the growth of high-quality training data, large language models must be trained with both compute and data constraints in mind. Unfortunately, existing prescriptions for data-constrained scaling suffer from two important limitations.
arXiv MLRead article →arXiv:2610.37833v1 Announce Type: new Abstract: Modern automatic speech recognition (ASR) systems have achieved remarkable accuracy by scaling model depth and capacity, but at the cost of increased memory, energy, and computational demands. On edge devices where ASR is often most needed, such as watches and glasses, deploying such large models is infeasible due to extremely constrained resources.
arXiv MLRead article →arXiv:2610.37828v1 Announce Type: new Abstract: An AI developer often has several candidate versions of a system and can deploy a mixture of them, sending each request to a randomly chosen version, as in A/B testing. Each stakeholder (e.g., a safety team, a regulator, or a customer with private test data) accepts a mixture exactly when its expected performance (e.g., its rate of harmful replies) meets the…
arXiv MLRead article →arXiv:2610.37823v1 Announce Type: new Abstract: Event cameras, spiking cochleas and similar sensors emit asynchronous events, yet most learning systems first bin them into frames or packets, so the output depends on how the stream was sliced. Networks whose state is updated exactly at every event and evolves in closed form between events are invariant to this slicing.
arXiv MLRead article →arXiv:2610.37819v1 Announce Type: new Abstract: Quantum machine learning (QML) is increasingly explored for medical classification, yet its behaviour under subgroup-aware fairness criteria remains understudied. We present an empirical fairness audit of classical and quantum classifiers for medical imaging, assuming QML is not inherently fair.
arXiv MLRead article →arXiv:2610.37796v1 Announce Type: new Abstract: Recent work has hypothesised that large language models may represent some concepts as multi-dimensional manifolds instead of single directions. Using these manifolds to interpret or steer representations often assumes that moving along them corresponds to a continuous change in the represented concept.
arXiv MLRead article →arXiv:2610.37737v1 Announce Type: new Abstract: Prior-Data Fitted Networks (PFNs) replace the hyperparameter refitting of Gaussian Process (GP) surrogates in Bayesian optimisation with a single forward pass, at the cost of a prior that is fixed at meta-training time.
arXiv MLRead article →arXiv:2610.37733v1 Announce Type: new Abstract: This work studies context-aware selective transfer learning for multi-well groundwater anomaly forecasting in the Netherlands. Source-domain pretraining can raise average accuracy but is not uniformly reliable: it degrades performance on a substantial fraction of heterogeneous wells.
arXiv MLRead article →arXiv:2610.37727v1 Announce Type: new Abstract: In machine learning, symbolic reasoning approaches often rely on static architectures that hinder integration into deep neural networks and gradient-based optimization. To address this, we introduce the Native Predicate Neuron, a differentiable logic-reasoning neural module that enriches standard architectures with symbolic reasoning capabilities via first-order logic…
arXiv MLRead article →arXiv:2610.37726v1 Announce Type: new Abstract: Accurate regression need not yield the correct equation after support selection. For a median/sample-standard-deviation threshold in neural PDE discovery, we derive an exact two-signal retention boundary and a complete-vector perturbation bound. Exact advection-diffusion libraries lose diffusion despite coefficient recovery to numerical precision.
arXiv MLRead article →arXiv:2610.37719v1 Announce Type: new Abstract: Post-training moves a language model's weights by about a thousandth of their norm and changes its behavior sharply; pretraining moves them by a third and buys a slow drift. We introduce leverage, an exact measure of what a unit of weight change buys: the Fisher quadratic form of an update, computed with one forward-mode pass per direction, divided by the same form for…
arXiv MLRead article →arXiv:2610.37712v1 Announce Type: new Abstract: Adaptive optimizers are widely used in deep learning, yet their behavior can degrade under heavy stochasticity and label noise where the sign and magnitude of gradients can be unstable.
arXiv MLRead article →arXiv:2610.38210v1 Announce Type: new Abstract: Weight-only, lossy quantization substantially reduces the storage and memory requirements for serving large language models (LLMs), but the growing key-value (KV) cache remains a bottleneck for long-context inference on personal devices.
arXiv AIRead article →arXiv:2610.38201v1 Announce Type: new Abstract: Designing power electronics converters traditionally requires intensive human expertise across topology sizing, commercial component selection, circuit simulation verification, and schematic drafting.
arXiv AIRead article →arXiv:2610.38194v1 Announce Type: new Abstract: Industrial shampoo-bottle surface defect detection must operate across many stock keeping units (SKUs), where differences in bottle geometry, material, label layout, reflectance, and imaging conditions can substantially alter the appearance of the same defect.
arXiv AIRead article →arXiv:2610.38175v1 Announce Type: new Abstract: Benders decomposition and column generation solve large optimization problems by passing cuts, columns, and dual values between a master problem and subproblems, and are hard to implement correctly. Large language models now automate parts of algorithm design, which points toward agents that build decomposition solvers for new problems.
arXiv AIRead article →arXiv:2610.38152v1 Announce Type: new Abstract: Lightweight representation bridges make pretrained, task-adapted EEG foundation models executable on other biosignal modalities without retraining their backbones. However, selecting a single bridge pathway globally assigns every instance to the same transfer route and ignores complementary errors across routes.
arXiv AIRead article →