Science & Mathematics

Low-fidelity (LF) optimization models are widely used in large-scale operations because they can be solved efficiently and reliably, but they often exhibit systematic errors relative to high-fidelity (HF) formulations that better capture physical and operational constraints. Existing learning-based approaches typically rely on supervised or semi-supervised training with HF solutions, which are expensive to compute and may not align with HF operational objectives. We propose a self-supervised framework that parameterizes LF optimization models directly from HF feedback. The low-fidelity problem is embedded as a differentiable optimization layer and trained end-to-end, while a HF correction problem evaluates its solution and provides gradient information through its dual variables, without requiring HF solution labels. We evaluate the framework on optimal power flow, optimal gas flow, and pooling problems, spanning different sources of low- and high-fidelity mismatch. Across all three applications, the self-supervised approach improves objective quality and feasibility over untuned LF models and generally outperforms semi-supervised training. The learned parameterizations also generalize to unseen operating conditions and, for network applications with up to 30,000 nodes, to systems not encountered during training.

Progress on foundation model pretraining is gated by who can afford a trial, and relational pretraining is no exception. Recipes are developed inside individual labs, reported under private protocols, and do not accumulate. Speedrun leaderboards remove that barrier by decentralizing the search rather than the training run. The problem is fixed, the stack is open, entries are ranked by wall-clock time to a quality target, and one GPU is enough to hold the record. We introduce modded-nanoRT, the first such leaderboard for Relational Transformers. Which database to hold out decides whether the benchmark can resolve improvements at all, so we fix it by calibration, ranking six candidates by how sharply a 45-minute learning curve separates improvement from seed noise. We open the leaderboard with typed trilinear attention, a relational operator that conditions attention scores and values on the foreign key traversed, together with Muon and SIGReg. The recipe reaches the target in 305.63 seconds on one NVIDIA H200, 8.11x faster than the reference implementation, and wins 13 of 15 pairs outside the development database. We will release the benchmark, its frozen artifacts, and the record runs. Our code is available at https://github.com/kunumi/nanoRT.

Few-shot image classification (FSIC) frequently suffers from visual ambiguity, where background clutter and RGB appearance similarities hinder discriminative feature matching in data-scarce settings. To mitigate this, we propose DAFIC, a geometry-guided framework that incorporates monocular depth into representation learning, episodic adaptation, and dense support-query matching. DAFIC first learns geometry-aware token representations through RGB-D self-supervised pre-training. Subsequently, the Depth-guided Geometric Adaptation (DGA) module uses depth-derived geometry to refine patch representations with geometry-enriched class context while suppressing background interference. Spatial Bidirectional Patch Matching (SBiM) then performs bidirectional dense matching on the refined features, reducing sensitivity to spurious correspondences and background clutter. Extensive experiments on four benchmarks show that DAFIC achieves strong performance against existing methods, with ablation and robustness analyses supporting the effectiveness of coupling geometry-conditioned episodic adaptation with bidirectional matching for few-shot visual recognition.

Distilled diffusion models generate high-quality images in a few sampling steps, but aligning them with task requirements and human preferences is difficult. Methods developed for multi-step diffusion models rely on differentiable rewards, or on score fields and stochastic trajectories that distilled generators may lack. We ask whether a few-step generator can be adapted using only external reward evaluations and reference samples, and show that two binary classifiers suffice. Our method, DualHeads, implements them as lightweight heads on a frozen diffusion backbone. The reward head learns which samples of the same prompt score higher, the reference head separates reference images from generated ones, and the generator maximizes a weighted sum of their logits. When the heads are optimal, this objective is KL-regularized maximization of the learned reward, and the ratio of head weights sets the regularization strength. Empirically, with both rule-based and model-based rewards, a few-step DualHeads generator reaches task rewards close to those of 40-step reinforcement learning methods while preserving image quality, at no extra inference cost. We hope that this simple and elegant dual-head formulation will help advance few-step reward-tilted generation.

Adoption of artificial intelligence in K-12 schools has outpaced guidance for students, teachers, and administrators. Existing AI-literacy frameworks emphasize understanding and use; cybersecurity frameworks predate AI-specific threats, and neither defines AI security literacy as a competency domain in its own right. We propose an AI Security Literacy (AISL) framework of six dimensions spanning the school ecosystem, ground its technical dimensions in public AI-safety benchmarks and a reproducible multi-seed adversarial evaluation, and ground its attitudinal dimensions in nationally representative survey micro-data from U.S. teachers and adults. Only 34.5% of teachers report a school AI policy, and teachers see their own AI use as easing their work but students' use as complicating it; on the technical side, a reported robustness figure depends as much on the attack and the particular trained model as on the system evaluated. NeuralFortify, a browser-based K-12 platform built on real AI models, is presented as an illustrative operationalization.

Jednym z najważniejszych wyzwań związanych z implementacją sieci Open RAN o architekturze zorientowanej na użytkownika jest odpowiednie formowanie tzw. klastrów obsługujących. Problem ten może być efektywnie rozwiązany za pomocą zaproponowanego Agenta AI bazującego na modelu językowym. Wyniki symulacji pokazały, że zaproponowane rozwiązanie pozwala na ponad 2-krotne zwiększenie mediany rozkładu przepływności użytkowników względem standardowej architektury zorientowanej na sieć.

A systematic understanding of the global structure of neural-network loss landscapes remains challenging due to their high dimensionality. Symmetries arising from network architectures and learning objectives provide a way to investigate aspects of this structure. Although a growing number of studies have drawn attention to these symmetries, a systematic characterization of them, as well as of the critical orbits they give rise to, remains incomplete. To address this gap, in this work we establish a topological symmetry index, based on the methodologies from the family of equivariant degree theories, to decode these symmetries. The frameworks apply to arbitrary network architectures and include methods applicable to degenerate critical orbits. The index detects the isotropy of the critical points and quantifies the corresponding net topological contributions. Several examples, including softplus MLPs, Transformers, and equivariant neural networks, are presented to illustrate the computations.

We design and optimize a quantum error correction (QEC) decoder for NVIDIA GPUs, combining Relay belief propagation, a state-of-the-art message-passing algorithm, with order-zero ordered statistics post-processing. Our implementation uses CUDA Graphs to control GPU execution, stops updates for converged samples, and collects unresolved samples in queues for post-processing. We optimize sparse traversal through coalesced message accesses, prefetching, and register reuse, and accelerate binary post-processing with warp-level collectives. In addition, mixed FP16-FP32 arithmetic and INT8 quantization reduce message storage and traffic while retaining accuracy using accumulators at higher precision. We evaluate surface, hypergraph product, and bivariate bicycle code workloads on NVIDIA A100 and GH200 data-center GPUs and a commodity RTX 5080. On the bivariate bicycle code at error probability 0.003, INT8 halves amortized decoding time relative to independently tuned FP32 with a negligible logical error rate increase on RTX 5080. With independent syndrome batches submitted back-to-back to concurrent decoder instances, RTX 5080 delivers 91% of GH200's highest measured throughput with an 8.9x shorter median submission-to-correction latency. The combination of performance optimization, mixed precision, INT8 quantization, and concurrent decoder instances opens a path toward QEC decoder clusters built from commodity GPUs.

Looped transformers reuse one block of layers many times and are often trained by backpropagating through only the last k loops. We measure how much of the weight gradient this throws away, exactly and for every k from one backward pass, on nine looped language models, and report three findings. First, on a trained model this lost gradient is set near the output: once the loop count is well above k, it barely changes with the loop count (within about 0.03 from 32 to 64 loops on the three models we tested, as a pre-registered test predicted for Parcae). Where the backward signal shrinks, $ρ^k$, with $ρ$ the signal's fitted shrink per loop, approximates it at moderate loop counts but overstates it at long ones, because the fitted $ρ$ drifts up as the fit takes in more loops. On Ouro the signal grows, and keeping two of four loops still loses 81% of the final output's gradient. Second, the lost gradient falls during fine-tuning, by 0.07-0.22 within 300 steps on three of four models, on held-out problems too. The fall is not just the read-out's share of the gradient; training only the layers outside the loop reproduced up to half of it, and it also falls under truncated training. So a trained model's value overstates what truncation removes later. Third, we could not show that it predicts the training cost of truncation better than rules that need no measurement: on two new models it tied Parcae's cost at the same k, in a test too small to separate them.

The weights of a trained network do not have uniform sensitivity to model accuracy: the 1% most sensitive weights carry about a fifth of the damage a bit flip can do. This distribution decides the bit-flip tolerance of a model, the fraction of weights a code must cover and the bits each weight needs; existing schemes work with the distribution training leaves. We show that we can systematically modulate or alter the weight sensitivity distribution of a trained model, thereby shifting weights from one sensitivity bin to another. The sensitivity of a stored weight is the product of its gradient and the step a flip takes in the storage format, and LEAN adds a composite regularizer on both factors, a range term that bounds the flip step and a tail term that lowers the gradient of the most sensitive weights, with one coefficient λ setting how far the distribution moves. Ten epochs with the term, as a fine-tune or as the last epochs of training, shrink the sensitive tail 5 to 13 times on nine vision models (ResNet-18, ResNet-50, VGG-16 on CIFAR-10, CIFAR-100, ImageNet) within 0.5 points of accuracy; real flips confirm that the reshaped weights do less damage, a plain gradient-norm penalty does not reproduce the effect, and a preliminary extension to three small language models shows the same tail shaping. With a BCH code over 1% of the weights the model then survives 300 progressive bit-search flips in every training seed and 500 in two of three, where three of four published defenses with the same code fail by 300 and WCAT falls to 75% at 500. The reshaped distribution also cuts the parity for a fixed guarantee 3.7 times and makes a sensitivity-ordered 8/6/4/3-bit format affordable, which stores ImageNet ResNet-50 at a 3.29-bit weight payload, 2.4 times smaller than INT8, above the accuracy of LSQ at 3 bits.

Purpose: To mitigate parameter degeneracy and improve robustness of IVIM fitting, a multimodal physics-informed neural network was developed to leverage anatomical information through the combined use of cross attention, selective spatial gating, and anatomy-weighted spatial prior. Methods: T1w and FLAIR anatomical images were co-registered to IVIM signal in native space and segmented by tissue type. The proposed Physics-Anchored Cross-attention Embedding network (PACE) encoded IVIM signals and their anatomical features in parallel branches, then integrated them via cross attention and spatial gating. An anatomy-weighted spatial prior further constrained signal reconstruction using anatomically similar neighboring voxels. PACE and its prior-guided variant were compared with two baseline deep neural networks (DNN and IVIM3C), least squares, non-negative least squares, and maximum a posteriori fitting models. All models were evaluated on signal and parameter reconstruction accuracy and lesion contrast fidelity using data generated from subjects drawn from the University of Rochester COVID cohort. Results: PACE variants demonstrated lower RMSE for all IVIM parameters except for Dmv. The anatomy-weighted spatial prior further reduced RMSE across all five parameters relative to the default PACE, with modest reductions in lesion contrast accuracy for fmv and fint. For in-vivo evaluation, PACE with spatial prior also showed the best signal reconstruction across all noise levels while maintaining good to excellent ICC. Conclusion: Multimodal anatomical information improved IVIM fitting accuracy and robustness, with selective spatial integration via cross attention, gating, and anatomy-weighted spatial prior. Keywords: intravoxel incoherent motion, physics-informed deep learning, diffusion MRI, multimodality fusion, attention mechanism

Large language models increasingly take part in extended conversations where they can both shape and be shaped by another party's views, yet most persuasion research measures only one side of that exchange. We introduce a controlled LLM-LLM debate framework in which two language-model roles repeatedly discuss a contested topic (climate-change attribution): one is given an assigned stance and, in some conditions, an explicit instruction to persuade, while the other receives no assigned position. Across four ordered model combinations involving GPT-6 Luna and Gemini 2.5 Flash Lite, we track both each role's explicitly reported position and a separate measure of its turn-by-turn behavior inside the conversation. Opinion dynamics differ substantially across model combinations: same-model interactions narrow the opinion gap on average, clearly for only one of the two models, while cross-model interactions widen it. Explicit reports and in-dialogue behavior capture distinct aspects of opinion dynamics, and a blanket instruction to persuade does not reliably increase influence. These findings suggest that conversational influence depends not only on assigned roles and initial positions but also on the participating models, with implications for how opinion change is measured in extended LLM interactions.

Ferroelectric layered semiconductors provide a unique platform for exploring coupled excitonic and electronic phenomena in reduced dimensions. Here, exciton transport in layered NbOI$_2$ is investigated using transient absorption microscopy. Ultrafast and strongly anisotropic exciton diffusion is observed. The diffusion coefficient of about 160 cm$^2$ s$^{-1}$ along the nonpolar crystal axis is among the highest reported for van der Waals semiconductors. The transport exhibits an approximately twofold in-plane anisotropy between the polar and nonpolar directions, reflecting the intrinsic low-symmetry crystal structure of NbOI$_2$. The environmental stability of NbOI$_2$ is further investigated, and fabrication and encapsulation conditions required to achieve long-term sample stability are identified. These results establish NbOI$_2$ as a promising platform for studying anisotropic exciton transport and highlight the critical role of interface protection in air-sensitive van der Waals materials.

We show that every class of distributions that is learnable up to arbitrarily small target error in total variation distance with finitely many samples is also learnable under ($\varepsilon, δ$)-differential privacy up to arbitrarily small target error with finitely many samples. This proves a conjecture of Ashtiani. Previous work by Bun, Kamath, Mouzakis, and Singhal [arXiv:2402.00267] gave a weak refutation of this conjecture, showing that this implication is not true when learnability is defined with respect to a fixed target error. In contrast, we show that, under the standard definition of learnability, the conjecture is true.

Human video scales easily and does not need robot hardware, but existing methods require depth sensing, overlay the robot onto 2D video, or generate data only for the demonstrated objects. We present Real2Sem2Real (R2Sem2R), a real-to-semantic-to-real pipeline that turns one monocular RGB video of a human demonstration into thousands of robot demonstrations across the semantic object instances of a category. R2Sem2R tracks the SE(3) pose of each object with DINOv3-distilled Gaussian splats, outperforming existing methods and FoundationPose with estimated metric depth. R2Sem2R then aligns digital cousins, each generated from one or a few photographs, to the demonstrated objects with a semantic alignment. This alignment matches whole objects and the functional regions of target objects, such as a coffee machine's tray. R2Sem2R maps the human hand to a feasible robot grasp that extends to the cousins. Robot demonstrations are generated either in IsaacSim, for photorealistic rendering, or MuJoCo, for contact physics, and R2Sem2R can generate over 2,000 episodes in about two hours. On two real robots and four tasks, we train a vision-language-action policy on data from one human video per task, and policies trained on our data match or exceed policies trained on teleoperation across diverse object instances. The project page is available at https://real2sem2real.github.io/.