Papers

Calderón preconditioning restores mesh-independent conditioning for boundary-element discretizations of the Maxwell electric field integral equation, but assembling the dual barycentric operator can dominate the computational cost. We show that this preconditioner can be assembled once at a nominal geometry and reused across connectivity-preserving shape perturbations. Exact combinatorial alignment of the RWG-BC degrees of freedom transports the duality pairing, leaving only the dual-mesh operator frozen. Under uniform discrete-stability and geometric operator-continuity assumptions, a bi-parametric operator-preconditioning analysis yields a spectral-condition-number bound uniform in the mesh size and, after tensorization, in the stochastic-Galerkin dimension. Numerical experiments on smooth and non-smooth scatterers show that the frozen preconditioner closely tracks freshly assembled Calderón preconditioners over substantial shape variations. A Monte Carlo campaign on a NASA almond, whose admissibility is screened before each shape is constructed so that no realization is rejected, separates the two strategies by less than the variation between realizations, and mesh refinement there leaves the frozen iteration count unchanged. Reuse yields roughly 30x measured speedup per new realization in our dense implementation; a cost-model estimate based on published compressed-assembly timings gives roughly 2x.

Agent memory systems are increasingly used to maintain long-term user preferences, task states and evolving facts, but current evaluations often collapse memory behavior into final-answer accuracy. We introduce MemProbe, a cognitive-science-inspired framework for diagnosing stability-plasticity tradeoffs in agent memory. The framework is motivated by a core insight from cognitive memory research: memory is reconstructive and shaped by interference, source reliability, reinforcement, and reactivation. MemProbe turns this insight into four reusable experimental paradigms (interference, misinformation, consolidation strength, and reconsolidation window) that manipulate when a memory should be updated, preserved, or treated as uncertain. It further decomposes correctness into behavioral profiles that reveal how systems update, preserve, attribute, and temporally organize information. We instantiate these paradigms in a 56-episode diagnostic suite and evaluate six incremental memory systems under a unified protocol. Results show that systems with similar aggregate scores exhibit distinct behavioral profiles. MemProbe provides such a diagnostic lens, turning aggregate performance into interpretable profiles of memory maintenance over time. Code is available at https://github.com/jq-ding/MemProbe.

Runtime failure monitors can use a model's internal representations to anticipate failures. We audit this monitoring strategy across two autonomous-driving tasks: online vectorized map generation with LaneSegNet and end-to-end planning with VAD. We find that frame-level errors are predictable at inference in both tasks. For LaneSegNet, a supervised latent probe reaches Area Under the Receiver Operating Characteristic curve (AUROC) 0.780 for high Chamfer error; to our knowledge, this is the first post-hoc frame-level failure monitor for online vectorized map generation. For VAD, a supervised planning-latent probe reaches AUROC 0.868 for mean-ADE failure. Our audit shows that internal access is not necessary for strong failure prediction. A monitor using only LaneSegNet's prediction outputs reaches AUROC 0.825, while for VAD, ego state, driving command, and the planner's predicted trajectory reach 0.924 on the same mean-ADE endpoint. Adding latent features to either baseline yields no statistically resolved improvement. This observation persists across a broad suite of planning failure endpoints, including endpoints whose labels depend on geometry unavailable to the non-latent baseline. Thus, predicting failure from an internal representation does not establish that the representation provides useful information beyond observable inputs and outputs. We propose an evaluation protocol for testing the incremental value of latent access and release our per-frame failure endpoint labels.

We study dynamic regret in online convex optimization with an indicator switching cost: a fixed penalty incurred whenever two consecutive decisions differ. This captures startup overheads such as server activation, model deployment, and cache updates, and on a bounded domain it recovers norm-based movement costs as a special case. Existing guarantees for indicator costs handle only static comparators. We show that a direct extension of these techniques to dynamic regret provably fails, motivating a different approach. We propose a meta-learning framework: a set of randomized lazy FTRL base learners restarted at dyadic time scales, aggregated by a movement-aware master that mixes their proposal densities and samples actions via maximal coupling of consecutive mixtures. The resulting algorithm satisfies, in expectation, $\mathcal{R}^{\mathbf{1}}_T \le \tilde{\mathcal{O}}(\min\{\sqrt{T(S_T{+}1)},T^{2/3}(P_T+1)^{1/3}\})$, where $\mathcal{R}^{\mathbf{1}}_T$ is the dynamic regret plus the cumulative indicator switching cost, $S_T$ counts comparator switches, and $P_T$ is the comparator path length. The bound holds simultaneously for all sequences and requires no prior knowledge of $S_T$ or $P_T$: it is minimax-optimal (up to logarithmic factors) for tracking piecewise-constant comparators, and also captures frequently moving comparators with small total path length.

We study proportional representation in temporal voting, where one candidate is selected in each round. While prior work has focused on approval ballots, we consider ranked preferences, which may change over time. A natural approach treats each voter's top candidates as approved, but the right cutoff may differ across voters and rounds. We therefore require proportionality to hold for every admissible choice of cutoffs, whether fixed and common, common but varying across rounds, or set individually for each voter in each round. Combining these interpretations with temporal versions of justified representation (JR), proportional JR (PJR), extended JR (EJR), and proportionality for solid coalitions (PSC) gives us a hierarchy of axioms. We ask which of these axioms can be guaranteed, and with how much knowledge of the future. Unlike with approval ballots, no version of EJR can be guaranteed, and for the other axioms, flexibility in the cutoffs comes at a price. With a fixed common cutoff, JR, PJR, and PSC can be guaranteed, but only by rules that see all preferences in advance. Once the cutoff may vary across rounds, even such rules cannot guarantee JR or PSC for groups that agree in only some rounds. For groups that agree in every round, however, knowing only the number of rounds suffices for PJR in polynomial time, and PSC needs no knowledge of the future at all. Under individual cutoffs, no version of JR or PJR can be guaranteed, yet a rule as simple as serial dictatorship achieves PJR up to an additive loss that no rule can improve on, however much it knows. Natural preference restrictions restore exact guarantees. Finally, we show that checking our axioms is often coNP-complete; but perhaps surprisingly, a stronger axiom can be easier to check.

Large language models (LLMs) have recently demonstrated promising capabilities as in-context policy optimizers for Reinforcement Learning (RL), enabling policy search driven by both numerical reward signals and natural language reasoning. However, deploying such methods in practice requires transmitting raw policy parameters and rewards history to cloud-based LLM APIs, exposing proprietary control strategies to third-party service providers. To address this issue, this paper introduces Privacy-Preserving Prompted Policy Search (PP-ProPS), a framework that enables LLM-guided policy optimization while keeping policy and environmental parameters confidential. PP-ProPS encodes policy parameters and reward values using secret client-side transformations before they are included in each API request, ensuring that the LLM provider observes only encoded policy parameters and scaled reward information. Furthermore, unlike Vanilla ProPS, the proposed framework does not require the true optimal episodic return to be known or disclosed to the LLM. Beyond protecting the optimization data, PP-ProPS improves the search process in two ways. First, it provides the LLM with individual reward components instead of only a single total return, offering more informative feedback about each candidate policy. Second, it uses a bounded history that prevents the prompt from growing indefinitely, improving search with high-dimensional policies and supporting the use of open-weight LLMs. The proposed PP-ProPS is evaluated on both continuous and discrete control problems spanning Multi-Joint dynamics with Contact (MuJoCo) locomotion, classic control, highway driving, and robotic arm manipulation. Compared to Vanilla ProPS, the proposed PP-ProPS outperforms ProPS in seven of the ten evaluated tasks, and surpasses conventional RL methods including PPO, SAC, and TRPO, in five of the six tasks.

Expensive evolutionary search does not always need an exact fitness estimate for every candidate. It often needs a reliable answer to a simpler question: which candidate is better? We address this need through Teacher-Guided Learning NSGA-II (TGL-NSGA-II), a low-fidelity framework for constrained Tiny Machine Learning (TinyML) neural architecture search. A pretrained teacher organizes samples into strata defined jointly by difficulty and class. Each candidate then undergoes KD-Lite, a short and capped knowledge-distillation procedure on a compact training set, before being scored on a separate stratified evaluation set. This teacher-guided score is fused with a Gaussian-process surrogate to select candidates for full evaluation. For a fixed candidate population, we analyse evaluation variance, score concentration, pairwise rank inversion, expected Kendall-$\tau$, first-front identification, and hypervolume perturbation. We also derive a variance-aware fusion weight and a capacity-adaptive distillation rule. On keyword spotting and bird-call classification, the measured Kendall-$\tau$ values are 0.74 and 0.62, exceeding the corresponding predicted lower bounds of 0.60 and 0.46. Joint stratification reduces proxy-score variance by 41% relative to random evaluation. Selective teacher mismatch, in contrast, increases differential bias and reduces Kendall-$\tau$ to 0.41. Under a constrained evaluation budget, TGL-NSGA-II achieves the largest mean hypervolume and smallest generational distance on keyword spotting, records the lowest mean false-positive rate on BirdCLEF, and runs 2.2x faster than full NSGA-II. These guarantees apply to population-level low-fidelity evaluation and do not establish convergence of the complete evolutionary trajectory.

As part of the JWST Hot Rocks Survey, we analyzed a JWST/MIRI F1500W secondary-eclipse observation of GJ 357 b, a rocky exoplanet orbiting a nearby M dwarf. We also performed a new global analysis of the GJ 357 system, combining TESS photometry, a JWST/NIRSpec transit observation, the MIRI secondary eclipse, radial velocities, and Gaia parallax measurements to obtain a self-consistent set of stellar and planetary parameters. We measure an occultation depth of $200.5 \pm 12.7,\mathrm{ppm}$, corresponding to a dayside brightness temperature at $15,\mu\mathrm{m}$ of $T_{\mathrm{b},15,\mu\mathrm{m}}=923^{+39}_{-38},\mathrm{K}$. The inferred brightness temperature substantially exceeds the zero-Bond-albedo, no-redistribution, disk-integrated bolometric dayside reference temperature of $686 \pm 14,\mathrm{K}$. The measured eclipse depth differs from the tested zero-albedo, grey airless-surface model by $5.2\sigma$ and is also inconsistent with the high-mean-molecular-weight atmospheric models considered here, including pure CO$_2$, pure H$_2$O, and N$_2$ with 100 ppm CO$_2$. Possible explanations for the excess emission in the MIRI/F1500W band include wavelength-dependent surface emissivity, localized hot regions, or atmospheric temperature structures and opacity sources not included in the tested models. Follow-up mid-infrared eclipse spectroscopy would constrain the shape of the dayside emission spectrum and help test the proposed surface and atmospheric explanations.

Resolved asymptotic giant branch (AGB) populations present many observational puzzles. We consider three often treated separately: photometric C/O classification boundaries must be redrawn empirically in new environments; hydrostatic model grids are too blue in bands sampling H2O absorption; and J-region (JAGB) luminosities can differ between fields of a single galaxy. Pulsation connects them into a coherent picture. Using SPHEREx spectroscopy of AGB stars in the Large Magellanic Cloud (LMC) with JWST/NIRCam photometry and HST time-series of the SN Ia host M101, placed on a common scale by synthetic photometry, we investigate how AGB variability, surface chemistry, and circumstellar dust depend on environment. We develop an F182M index sensitive to 1.9$\mu$m H2O absorption that chemically classifies variable AGB stars and reveals oxygen-rich (O-rich) stars degenerate with carbon-rich (C-rich) stars in broadband colors. Because these features arise in extended atmospheric layers, hydrostatic grids struggle to reproduce them. Pulsation and circumstellar dust are asymmetrically linked: nearly all stars with substantial infrared excess are variable, while many variables show little dust. This supports mass loss initiated by pulsation and enhanced by dust formation. Variable O-rich AGB stars in M101 are up to ~0.7 mag redder than their LMC counterparts in dust-sensitive colors, consistent with more efficient silicate-dust production at higher metallicity, while C-rich populations depend less on environment. Color windows that select C-rich stars cleanly in the LMC are therefore 33-68% O-rich in the metal-rich M101. This challenges the J-region's universality as a standard candle while evading standard self-consistency checks. We recommend combining variability, or the infrared excess that implies it, with photometry of molecular features to improve AGB classification at 0.75-5.0$\mu$m.

Benchy is a semantic language and execution engine for benchmarking AI programs. A benchmark is completely specified by a program, a scoring function, and a dataset, B=(P,S,D), and is separate from the AI-system taking it; a run binds the two, R=(B,AI). Benchmarks are authored as canonical YAML in which each semantic concept has one valid syntax, classified by a shared task/domain/language ontology, and deterministically compiled into a canonical JSON intermediate representation that the engine executes. Compilation changes representation, not meaning: it does not repair invalid definitions or inject hidden defaults. Programs use fixed schemas of named input and output fields, the leaf output fields are the scoring dimensions, and the engine exposes one universal runtime contract — a named-field input object in, a named-field output object out — to which external AI-systems adapt at the boundary, so integration mechanics never propagate into benchmark semantics. This paper gives the semantic object model, the ontology and task-to-program validation rule, the scoring and failure semantics, the compilation and execution architecture, and the scope of the current language. An appendix fixes the normative engineering contract for the first engine implementation.

We study the structural, thermophysical and dynamical properties of molten MgCl2 over a wide temperature range using Ab Initio Molecular Dynamics (AIMD) and molecular dynamics simulations based on machine-learning interaction potentials (MLIPs). MLIPs can achieve near-AIMD accuracy at a fraction of the computational cost, enabling simulations of larger systems and longer timescales. This opens up the possibility of studying the system under out-of-equilibrium conditions, which in turn allows for the calculation of physical properties, such as the thermal conductivity or viscosity, that cannot be obtained reliably for the typical time and length scales accessible to AIMD. We follow two complementary approaches to develop and evaluate the MLIPs. Firstly, we develop a Deep Potential (DP) from scratch using AIMD trajectories at different temperatures. Secondly, we also evaluate two out-of-the-box foundation models and a fine-tuned version of one of them. The fine-tuning is performed using configurations from the AIMD trajectories originally used to train the Deep Potential. We compare the predictions of the trained DP, the foundation models and AIMD simulations with the available experimental data and provide a comprehensive calculation of the most important physical quantities of molten MgCl2. We compute densities, radial and angular distribution functions, thermal conductivity, heat capacity, viscosity and diffusion coefficients over the 1000-1600 K temperature range. This range is relevant to important technological applications such as thermal energy storage, generation IV nuclear reactors and concentrated solar power systems. Finally, we briefly review the accuracy and performance of the different MLIPs versions used throughout the work.

Instrument timbre transfer re-voices a performance using the timbre of another instrument. Extracting the target timbre from an audio reference capture more nuances than inferring it from a text prompt. Systems that read timbre from such a clip train a dedicated model for the task, which captures the timbre cleanly but stays a narrow, single-purpose system. More versatile approaches add control to a pretrained music generator, yet a reference clip entangles timbre with genre and melody, so these systems fall back on text to name the timbre. We present MuseTimbre, the first system, to our best knowledge, that transfers timbre from an audio reference to a polyphonic source through conditioning a pretrained music generator. This system employs a multi-pitch estimator to extract pitch information from the source and finetune a CLAP encoder to extract timbre information from the reference audio. Experiments show that across four datasets of real polyphonic recordings, MuseTimbre achieves pitch alignment on par with the baselines while matching the reference timbre far more closely. Results also show that the finetuned CLAP-based timbre extractor is robust to pitch variations, making it useful in timbre similarity measures.

Audience sizing is a critical component of digital marketing. It enables precise resource allocation, campaign planning, and performance optimization. Traditional approaches using skeleton audiences, sampling, or predictive modeling suffer from significant delays, estimation errors, and poor scalability over high-dimensional profile data. We present REALMS (Real-time Exact Audience sizing via LLM-based Multi-attribute Search), a conversational system for exact audience sizing deployed in production on an enterprise customer data platform. REALMS enables marketers to query massive profile stores with millions of profiles and thousands of attributes using natural language and receive precise counts in seconds. The system introduces three key components: (1) a categorical attribute retrieval mechanism using embedding-based vector search to dynamically identify relevant schema attributes without manual configuration; (2) an LLM-powered NL2SQL pipeline with template-based in-context learning for accurate query generation over complex nested schemas; and (3) schema standardization enabling industry-agnostic deployment across diverse enterprise environments. Evaluation on real enterprise data demonstrates strong recall for attribute retrieval, high SQL execution accuracy, and low latency, which enables real-time interactive audience insights where prior methods required hours.

In this paper, we establish the first asymptotic convergence guarantees for the Muon algorithm through a more accurate proxy for the Newton-Schultz iteration than the typical matrix sign function. We prove that, for appropriate choices of hyperparameters, the iterates satisfy $\lim_{k\to\infty}\|\nabla f(x_k)\|=0$, and, under a global Polyak-Łojasiewicz condition, that the sequence of function values converges linearly. The key insight is that the regularization, implicit in Muon's Newton-Schulz implementation, induces a bounded preconditioner, exposing Muon as a preconditioned Polyak heavy-ball method and enabling a classical Lyapunov analysis. This observation naturally motivates applying the same preconditioning structure to the Nesterov gradient evaluation. We formalize this idea by introducing Muesterov, a Nesterov-based variant of Muon, and prove that it enjoys the same convergence guarantees, extending the theoretical framework beyond the heavy-ball setting. Numerical experiments on a scalar cross-entropy problem corroborate the theory and illuminate the joint role of the learning rate and the Newton-Schulz regularizer in controlling convergence. Preliminary numerical simulations training the nanoGPT dataset provide intuition regarding the relevance of the observations in this paper to practical applications.

The Weyssenhoff fluid expands the ordinary perfect-fluid description of general relativity by taking into account the intrinsic spin of its constituent particles. Within the linearized regime of Einstein-Cartan theory, we investigate the exterior geometry produced by isolated slowly rotating Weyssenhoff fluids, i.e., sources endowed with both macroscopic and microscopic angular momenta. Our work represents an extension of the foundational analysis of Arkuszewski, Kopczyński and Ponomariev, which was restricted to the static fluid configuration. Under the assumptions of stationarity and axial symmetry, we derive the explicit expression for the linearized stress-energy tensor of the source and obtain the corresponding metric through a multipole expansion. The ensuing solution exhibits a generalized Lense-Thirring structure, with both the gravitoelectric and gravitomagnetic sectors affected by the quantum spin and macroscopic rotation. We also discuss some phenomenological implications of the model, including the gyroscope precession and the Sagnac effect, along with potential experimental probes of the predicted corrections.

We study KMS and ground states of the tame $C^*$-algebras $\mathcal{O}_{m,n}$ of Ara, Exel and Katsura under the dynamics $\sigma_t(s_i)=m^{it}s_i$, $\sigma_t(t_j)=n^{it}t_j$. Normalized restriction to the full corner $A=q\mathcal{O}_{m,n}q$, generated by the partial isometries $v_{ij}=s_i^{*}t_j$ of common energy $\log(n/m)$, is at each inverse temperature $\beta$ an affine weak-$*$ homeomorphism of the two systems. Let $m,n\ge 2$. When $m=n$ the corner dynamics is trivial, so the KMS simplices are identified with trace simplices. When $m\neq n$, KMS states occur only at $\beta=1$, and there they are not unique. Conformal branching models surject the KMS$_1$ simplex onto a transportation polytope of dimension $(m-1)(n-1)$, whose vertices lift to extremal, hence factorial, KMS$_1$ states, the fibre over each vertex being itself the KMS$_1$ simplex of a quotient of $A$.

As robots transition from structured factory settings into homes, they are required to interact with an ever-increasing variety of objects. Many tasks require grasping, and often it is not sufficient to just pick up the target object. Consider a task like "pouring coffee" — to facilitate the subsequent pouring, the robot should grasp the mug by its handle. Existing learning-based approaches for grasping either find robust and collision-free grasps that are largely agnostic to the task (e.g., picking up the mug by its rim), or leverage foundation models to propose task-appropriate grasp locations that lack fine-grained physical grounding (e.g., reaching for and missing the handle). In this work, we bridge these approaches with a real-to-sim-to-real framework. Based on a single RGB-D observation, we construct a digital twin of the environment, query a large foundation model to propose grasps that align with the object's affordances and task description, and then optimize the proposals to ensure robustness and plausibility before executing the result on the real robot. Our key insight is that the grasp proposals of the foundation model should be regarded as semantic priors that serve as seeds for local, gradient-free optimization. We leverage Bayesian optimization with Thompson sampling to draw batches of nearby poses, which are subsequently evaluated in parallel under domain-randomized physics rollouts. The resulting grasp is both task-oriented and physically feasible for execution by the robot arm. Our full zero-shot real-world transfer only takes a few minutes and improves task-oriented grasping success by up to 33% as compared to other state-of-the-art pipelines. Our code is available here: https://github.com/VT-Collab/GraspTwin/

De novo peptide sequencing from tandem mass spectra is essential for identifying peptides without relying on reference databases. Despite advances in deep learning, accurate sequencing remains challenging because experimental spectra are often sparse, noisy, and incomplete, leaving informative b- and y-ion fragments unobserved. Existing methods attempt to recover this missing evidence via latent-space imputation before autoregressive decoding. However, they typically treat imputation as a fixed reconstruction task, without considering which missing fragments are most relevant to decoder errors. Moreover, existing peak representations do not explicitly model mass differences between peaks, despite their fundamental importance. We introduce GyroNovo, a framework with two main contributions. First, we use decoder errors observed during training to adapt the imputation objective, prioritizing fragments associated with frequent decoding errors. We further use the decoder error distribution to construct easy and hard augmented views of each spectrum, enabling the decoder to learn under varying degrees of spectral corruption and missing-fragment severity. Second, we introduce a mass-aware inductive bias into self-attention by using rotary embeddings to encode pairwise mass differences between spectral peaks. Together, these components align missing-fragment recovery with decoder behavior while explicitly incorporating the mass relationships that underlie peptide fragmentation. At inference time, GyroNovo retains a standard encoder-imputer-decoder architecture and requires neither additional inputs nor auxiliary search procedures. Experiments on NovoBench show gains of about 9 percentage points in peptide-level precision and 7 percentage points in amino-acid-level precision over the state-of-the-art baseline. Code: https://github.com/UBC-NLP/gyronovo.

Optimizing embedding systems for production recommendation pipelines demands systematic exploration that consumes disproportionate engineering effort at scale. We apply Andrej Karpathy's AutoResearch paradigm – a large language model that iteratively edits a training script and retains modifications that improve a held-out scalar metric – to automate this exploration. We report on twelve weeks of running this paradigm at production scale, where iterations consume hours of multi-GPU compute, evaluation involves competing criteria, and campaigns span weeks across many training jobs. Across two independently developed representation-learning systems for a book recommendation pipeline, we ran 220+ experiments and observed five recurring failure modes absent from the original setting: infrastructure fragility, agent memory decay, search-direction stagnation, iteration-cost asymmetry, and metric fixation. We contribute a three-principle scaffolding design – prevent, persist, redirect – that maps each failure mode to a structural remedy and whose instantiation scales with iteration cost. The framework produced a 1.82x Recall@6 lift and a 2.1x coherence lift over hand-tuned baselines, and the agent autonomously designed a text-only fallback that expanded catalog coverage by 5.8x. The two systems span nearly three orders of magnitude in per-iteration cost yet exhibit the same failure modes, suggesting these are structural properties of production-scale autonomous research rather than artifacts of either application.

Text-to-music systems are assessed on audio quality and on how faithfully the music follows its prompt, and the CLAP score, the cosine similarity between a music-text model's audio and text embeddings, is the standard objective metric of faithfulness. We ask how accurately that score reflects the text: when an attribute is linked to an instrument (e.g., distorted guitar), does the text embedding capture that binding? To find out, we introduce an attribute swap perturbation: the caption of a real recording is edited by exchanging exactly one property, timbre, lead versus accompaniment, or order of first appearance, between two instruments. We then test four contrastive music-text models and one large audio-language model on whether the audio scores higher against the original caption than against the perturbed one. No contrastive model distinguishes the two captions reliably. The audio-language model does better, but further experiments show that its advantage rests largely on audio-agnostic language priors. Our results thus provide compelling evidence that the CLAP score and related metrics do not capture fine-grained musical meaning or attribute bindings; their representation is closer to a bag-of-words that leaves them insensitive to meaning-changing perturbations of the caption.

Split Gibbs sampling enables diffusion posterior inference for general nonlinear inverse problems by decoupling prior and likelihood computations, allowing a pretrained diffusion prior to be reused across measurement models. However, its likelihood update often relies on iterative MCMC, which can hinder parallelization, require algorithm-specific tuning, and incur substantial computational cost. In this work, we propose a learning-based framework to replace this MCMC step by reformulating both Gibbs updates as Gaussian denoising problems and implementing them through ODE diffusion. The prior step reuses a pretrained denoiser, while the likelihood denoiser exploits known likelihood structure through a lightweight deep-unfolded network. Experiments on nonlinear phase retrieval demonstrate the effectiveness of the proposed method as an alternative to MCMC-based split Gibbs at lower likelihood-update cost.

SRAM-based FPGA system-on-chip devices are vulnerable to single-event upsets (SEUs) in configuration memory, which may perturb programmable routing resources and degrade communication fabrics. In modern Zynq UltraScale+ MPSoCs, such routing disturbances can introduce small propagation delay shifts that remain logically transparent yet compromise AXI-based data transfers and lead to silent data corruption. Although routing delay degradation and AXI interconnect failures have been studied independently, their experimental correlation under neutron irradiation has not been established. This work presents a cross-layer investigation on a ZCU104 platform integrating routing-dominated delay sensors with an AXI interconnect benchmark comprising replicated accelerators. Neutron irradiation experiments were conducted on the fully operational system, while a frame-level configuration fault injector implemented via the internal configuration access port enables controlled upset emulation. Measured routing delay events are statistically correlated with communication failures, and cross-sections for both timing shifts and AXI malfunctions are derived. The results experimentally demonstrate how neutron-induced routing perturbations propagate into system-level silent data corruption in UltraScale+ MPSoCs, providing insight for resilience-oriented AXI-based design in neutron-rich environments.

Using recent advances in the integral canonical models of Shimura varieties using syntomic methods, we give an extension of the Langlands–Kottwitz–Scholze method from Scholze's paper on the Langlands–Kottwitz method for deformation spaces of $p$-divisible groups. Namely, we define analogues $\phi^{\mathcal{G},\mu}_{\tau,h}$ of the local test functions from op. cit. in full generality, show they satisfy reasonable harmonic-analytic properties, and give rise to trace formulae for the Galois-Hecke action on the cohomology of Shimura varieties of bad reduction. Additionally, we use these new local test functions and the work of Fargues–Scholze to state a more robust and unconditional version of the Scholze–Shin conjecture. Along the way we study the $(\mathcal{G},\mu)$-apertures of Drinfeld and Gardner–Madapusi, in particular showing that such $(\mathcal{G},\mu)$-apertures carry no more information than a Tannakian version of Fontaine–Laffaille theory, at least over nice bases, and give a very explicit description of the universal deformations of $(\mathcal{G},\mu)$-apertures.

Astronomy papers carry much of their content in figures, but literature search indexes text, so there is no way to find a plot by describing what it shows. We present FOTO, a figure-level search tool over 482,750 captions from 53,839 peer-reviewed astro-ph papers. Each figure is indexed once by its title and caption, embedded with a 110M-parameter open model that runs locally, and a large language model is applied only to the retrieved shortlist, where it verifies real figures rather than generating references to ones that do not exist. Held to the harder figure-level criterion, FOTO recovers the target for 29 to 79% of queries at recall 20 depending on register, against 6 to 20% for Pathfinder and 8 to 16% for Semantic Scholar on the easier paper-level criterion, and it beats the paid embedding API it replaced. Reranking the shortlist then raises recall at 1 from 0.667 to 0.875 on detailed queries and from 0.071 to 0.571 on vague ones. Openly available models match or beat paid LLM models for verification

Children in multilingual communities often code-switch, using multiple languages in a single utterance. Can we induce cross-lingual alignment in language models by training on code-switched text? We pretrain small decoder-only transformers on two 100M-word multilingual corpora: a base corpus formed by mixing the English, Dutch, and Chinese BabyBabelLM datasets, and a corpus generated from it by inserting word- and sentence-level code-switching using an LLM. We find that training on code-switched data aligns the representations of parallel text, particularly across different scripts, and that this alignment persists through training on monolingual documents. Under a learning curriculum that progresses from word-level code-switching, to sentence-level code-switching, to monolingual documents, models trained on code-switched data outperform baselines trained without it on the BabyLM evaluation suite. Our work characterizes code-switching curriculum learning as an effective data augmentation method for multilingual pretraining. We release our code, data, and models at https://github.com/drooryck/multilingual-macaroni.

Configuration-induced perturbations in SRAM-based FPGAs may activate dormant programmable routing branches and increase path delay without immediately producing a functional error. Although prior studies have separately investigated the electrical origin of these delay changes, their in-situ detection, and the topology of commercial routing fabrics, a scalable method for predicting their timing impact at the granularity of programmable interconnect points and routed nets remains unavailable. This paper presents GRACIDIT, a graph-circuit digital twin framework for predicting configuration-induced routing delay degradation in Zynq UltraScale+ FPGAs. The proposed framework extracts the routing-resource graph of the XCZU7EV programmable fabric from the vendor design database, identifies inactive programmable interconnect points adjacent to active routes, and represents each candidate perturbation through its branch topology, geometric span, fan-out, physical region, and downstream loading. These graph features are combined with a calibrated reduced-order electrical model to estimate the delay introduced by single and cumulative routing-branch activations. Controlled configuration-equivalent perturbations are generated on a ZCU104 platform and characterized using complementary routing-domain oscillators and phase-sweep probes. The resulting model associates predicted delay shifts with available timing slack to rank vulnerable programmable interconnect points and routed nets and to construct a spatial vulnerability atlas of the programmable fabric. Experimental evaluation demonstrates a mean absolute prediction error of 7.8 ps, achieves 87.4 percent recall for slack-violating perturbations, and attains a Recall at 10 value of 0.90 for the most vulnerable routing resources.

We propose ST-pRRTC, a GPU-parallel space- time RRT-Connect motion planner for problems with known obstacle trajectories and unspecified arrival time. Searching over many arrival times broadens temporal coverage but divides a finite planning budget among more backward trees. To address the challenge, ST-pRRTC builds a shared forward tree and an adaptive forest of backward goal-time trees. Its interval root formulation samples goal arrival times continuously and guarantees probabilistic completeness and asymptotic arrival- time optimality under the stated assumptions in a bounded time domain. The practical root recycling policy has no such guar- antees. It adapts a fixed number of backward trees, replacing later roots while retaining useful search progress. Experiments on three dynamic benchmarks show that both variants achieve lower mean first-solution times and earlier mean final arrivals than ST-RRT* and SI-RRT on problems solved by all compared methods. Further experiments demonstrate the benefit of recy- cling over broad arrival-time ranges. Real-robot demonstrations show root-recycling ST-pRRTC planning motions for a UR5e among moving Crazyflie quadrotors.

Today's Quantum Processing Units (QPUs) are too small and too noisy to solve large combinatorial optimization problems directly, so practical hybrid solvers split a problem into pieces and iterate a decompose-solve-aggregate loop over whatever backends are available: classical heuristics, simulators, emulators, or a QPU. In practice, the loop is a driver script. It sits on top of the quantum-HPC middleware, handling task generation, provenance, recovery, and portability. Instead, we treat the loop as a scientific workflow and ask what a workflow layer adds to a generic workflow management system and QPU-sharing middleware. Two decomposition patterns from real applications, iterative consensus (ADMM) and hierarchical partitioning, turn out to stress the orchestration layer very differently: over 120 managed runs, orchestration took 78.4% of end-to-end time for the iterative pattern, almost all of it in a per-round barrier, but only 6.2% for the hierarchical one. Our workflow model adds four things a generic engine does not have: a termination predicate residing in the task graph that reads the previous round's residuals, subproblem-level recovery with warm-start and quorum-deferred aggregation, failover from a QPU to a classical replica within a round, and a provenance schema that describes CPUs, simulators and QPUs with the same fields, including the shot budget included. We report the cost of each layer on our engine, show that speculative re-execution enable runs to complete under injected failures that stall an unmanaged driver. We also use the same provenance to give per-device latency tails across simulators, emulators and IQM QPUs.

We classify all real extremals of the sharp $L^2$ Folland-Stein inequality on H-type groups and prove a global Bianchi-Egnell stability estimate. The linearized kernel at the standard bubble consists exactly of its translation and dilation derivatives. We also obtain explicit bounds for the spectral gap above the symmetry modes and characterize the optimal coefficient in the local stability asymptotics. The results apply to arbitrary orthogonal Clifford modules.

We demonstrate a jumping robot that reaches high (7.6 m) and fast (190 ms stance time) jumps from a single long leg driven by a direct-drive transmission, without elastic energy storage. At 281 g, it achieves the highest jump yet reported for an electrically actuated, spring-free system. The leg uses a new fabric-wrap transmission that provides a variable mechanical advantage, keeping a small electric motor near its peak power output through the stroke while bracing the long, lightweight leg against buckling. A balancing module at the top of the leg uses small propellers to control the leg's orientation on the ground and in the air, where the long leg provides a large moment arm for the control torques. The robot is validated outdoors with vertical jumps, attitude control on the ground and in flight, and tilted jumps.